Full text of "2001 -The Quest For The Origins Of Vedic Culture The Indo Aryan Migration Debate"

See other formats

the Quest for
the Origins of
Vedic Culture

The Indo-Aryan
Migration Debate

EDWIN BRYANT

The Quest for the Origins of Vedic Culture

This page intentionally left blank

The Quest for the Origins
of Vedic Culture

The Indo'Aryan Migration Debate

EDWIN BRYANT

OXFORD

UNIVERSITY PRESS
2001

OXFORD

UNIVERSITY PRESS

Oxford New York

Athens Auckland Bangkok Bogota Bombay Buenos Aires Cape Town
Chennai Dar es Salaam Delhi Florence Hong Kong Istanbul Karachi
Kolkata Kuala Lumpur Madrid Melbourne Mexico City Mumbai
Nairobi Paris Sao Paulo Shanghai Singapore Taipei Tokyo Toronto Warsaw

and associated companies in
Berlin Ibadan

Published by Oxford University Press, Inc.

198 Madison Avenue, New York, New York 10016

Oxford is a registered trademark of Oxford University Press

All rights reserved. No part of this publication may be reproduced,
stored in a retrieval system, or transmitted, in any form or by any means,
electronic, mechanical photocopying, recording, or otherwise,
without the prior permission of Oxford University Press.

Library of Congress Cataloging-in-Publication Data
Bryant, Edwin

The quest for the origins of Vedic culture :
the Indo-Aryan migration debate / Edwin Bryant,
p. cm.

Includes biblioeranhicnl reference and index.

ISBN 0-19-513777-9; ISBN 0-19-516947-6 (PBK)

1. Indo-Aryans—Origin. 2. India—History—To 324 B.C.

3. Indus civilization I. Title.

DS425.B79 2000
934'.02-dc21 994386274

First published as an Oxford University Press paperback 2003

1 35798642

Printed in the United States of America
on acid-free paper

To my father and sister
for all their support.
And to Fran, Ted, and Jack
for making this possible.

This page intentionally left blank

Acknowledgments

To Fran Pritchet, for taking me under her wing right from the start; Ted Riccardi, for
always encouraging me to pursue my intellecmal interests; and Jack Hawley, for making
sure that I came up to standard. To Gary Tubb, for his comments, meticulous as al¬
ways, and to Michael Witzel, for being so generous with his time and vast learning. To
James Mallory, for extensive comments, and to Kim Plofker, for valuable criticisms. To
Richard Meadow, Hans H. Hock, Hermut Scharfe, Peter Rahul Das, Jay Jasanoff, and
Thomas Trautmann for providing feedback on various chapters or sections of this work.
To Fred Smith, Carl Lamberg-Karlovsky, and Vasudha Narayanan for their help. To
the many scholars whom I had the good fortune to meet in India, and who were so
generous with their time and hospitality. Needless to say, the views represented herein
are my own and not necessarily of those who have been kind enough to point out the
most egregious errors in previous drafts of this work. Special thanks to Matthew Ekstrand-
Abueg for doing a great job with the maps and diagrams despite the last-minute time
constraints, and to Martin von Wyss from the Harvard map room; to Fritz Staal for his
assistance in obtaining the photograph for the cover; to Pia Bryant for her extensive
editing; to the American Institute of Indian Studies for providing me with a research
and travel grant, and to the Charlotte Newcombe Foundation for offering me a Ph.D
write-up grant.

This page intentionally left blank

Contents

Introduction, 3

1. Myths of Origin: Europe and the Aryan Homeland Quest, 13

Biblical Origins, 14

India, the Cradle of Civilization, 18

The Aryans and Colonial and Missionary Discourse, 21

German Aryanism, 30

Two Centuries of Homeland Theories, 35

Present-Day Homeland Hypotheses, 38

Conclusion, 43

2. Early Indian Responses, 46

Hindu Nationalist Responses, 47

The First Reactions: Hindu Religious Leaders, 51

Conclusion, 56

3. Vedic Philology, 57

The Racial Evidence, 59

The West-to-East Geographic Shift in Sanskrit Texts, 63
Conclusion, 67

4. Indo-European Comparative Linguistics: The Dethronement of Sanskrit, 68

The Law of Palatals and the Discovery of Hittite, 69
Objections from India, 72
Conclusion, 73

5. Linguistic Substrata in Sanskrit Texts, 76

Linguistic Innovations in Sanskrit, 78

Evidence of the Loanwords, 84

Terms for Flora in Indie Languages, 90

Place-Names and River Names, 98

Indo-Aryan, or Dravidian and Munda Migrations?, 102

Conclusion, 105

6. Linguistic Paleontology, 108

Flora and Fauna, 109
The Horse, 115

Contents

Criticisms of the Method, 120
Conclusion, 123

7. Linguistic Evidence from outside of India, 124

Semitic Loans in Indo-European: Nichols’s Model, 124

Finno-Ugric Loans, 126

Other Traces of Indo-Aryan, 129

The Avestan Evidence, 130

The Mitanni Treaties, 135

Conclusion, 138

8. The Viability of a South Asian Homeland, 140

Center of Origin Method, 142

Dialectical Subgroupings: Gamkrelidze and Ivanov’s Model, 145
Nichols’s Sogdiana Model, 151
Conclusion, 154

9. The Indus Valley Civilization, 157

Indra Stands Accused, 157

The Religion of the Indus Valley, 160

The Sarasvatl, 165

The Horse, 169

The Chariot, 1 75

The Indus Script, 1 77

Urbanity and the Rgveda, 184

Conclusion, 192

10. Aryans in the Archaeological Record: The Evidence outside the Subcontinent, 197

Identifying Aryans, 198
The Northern Route, 202
The Southern Route, 208
Two Wave Theories, 217
Conclusion, 220

11. Aryans in the Archaeological Record: The Evidence inside the Subcontinent, 224

Gandhara Grave Culture, 225
Jhukar Culture, 226
Cemetery H Culture, 229
Painted Gray Ware Culture, 229
Aryans in the Skeletal Record, 230
Continuity and Innovation, 231
Conclusion, 236

12. The Date of the Veda, 238

Dating Proto-Indo-European, 239
Dating the Veda, 243
Astronomy and Vedic Chronology, 251
The Mathematics of the Sulvasautras, 262
Conclusion, 264

Contents

13. Aryan Origins and Modern Nationalist Discourse, 267

Nationalism and Historiography: General Comments, 268
The Aryans in Hindutva Ideology, 270
Stereotypes and Counterstereotypes, 275
Discourses of Suspicion, 286
Conclusion, 295

Conclusion, 298

Notes, 311

Works Cited, 349

Index, 381

This page intentionally left blank

The Quest for the Origins of Vedic Culture

This page intentionally left blank

Introduction

The solution to the Indo-European problem has been one of the most consuming intel¬
lectual projects of the last two centuries. It has captivated the imagination and dedica¬
tion of generations of archaeologists, linguists, philologists, anthropologists, historians,
and all manner of scholarly, and not so scholarly, dilettantes. Predicated on the deduc¬
tion that cognate languages necessitate an original protoform spoken by a group of people
inhabiting a reasonably delineated geographic area, the problem has resulted in a mas¬
sive amount of scholarship attempting to reconstruct this protolanguage, locate the original
homeland where it was spoken, and conjecture on the social and cultural life of the
protospeakers. Although the endeavor has very much been a preoccupation of European
scholars, the belief in and pursuit of the origins of European civilization have required
scholars to attempt to reconstruct and reconfigure the prehistory and protohistories of
other civilizations whose languages happen to belong to the Indo-European language
family.

The publicization, in Europe, of the Sanskrit language and of its connection with
the classical languages of Europe was the catalyst for the whole post-Enlightenment quest
for the Indo-Europeans that continues, unresolved, to this day. This “discovery” of
Sanskrit resulted in the earliest history of the Indian subcontinent also being subsumed
by the problem of European origins. Although India was initially entertained as the
homeland of all the Indo-Europeans, various arguments were raised against this pro¬
posal, and Indian civilization was construed as the joint product of an invading Indo-
European people—the Indo-Aryan branch of the family—and indigenous non-Indo-
European peoples. Yet although taking it upon themselves to determine the history of
the Indian subcontinent in accordance with the currents of scholarship that have ebbed
and flowed in academic circles in Europe over the decades, Western scholars have gen¬
erally been unaware, or dismissive, of voices from India itself that have been critical
over die years of this European reconstruction of their country’s history. In the words
of one of the scholars who will be featured here, “However well-meaning such [schol¬
ars] . . . and their publications are, they have taken it upon themselves the task of inter¬
preting the past heritage of a very large number of people who belong to various nation
states and may like to formulate their own ideas of the past” (Chakrabarti 1997, 207).

4 Introduction

This book is primarily a historiographical study of how various Indian scholars, over
the course of a century or more, have rejected this idea of an external origin of the Indo-
Aryans by questioning much of the logic, assumptions, and methods upon which the
theory is based. The aim of the book is threefold. A primary aim is to excavate mar-
ginalized points of view reacting against what is perceived as a flawed and biased his¬
torical construct. As a corollary of this aim, this work will further complicate the Indo-
European homeland quest by exposing the whole endeavor to a critique from scholars
outside mainstream European academic circles who do not share the same intellectual
history as their Western peers. A further aim of this book is to present a comprehensive
exposition and analysis of views from within mainstream academic circles addressing
the issue of Indo-Aryan origins.

With regard to the primary aim, I have used the term Indigenous Aryanism to denote
a theme that is common to many of the scholars I examine in this book. This runs the
risk of essentializing a quite variegated cast of characters, and I merely use the term to
encompass a position on Indian protohistory that I view as common to most of the
arguments that I examine. The scholars referred to by this term all share a conviction
that the theory of an external origin of the Indo-Aryan speaking people on the Indian
subcontinent has been constructed on flimsy or false assumptions and conjectures. As
far as such scholars are concerned, no compelling evidence has yet been produced to
posit an external origin of the Indo-Aryans.

The various scholars whose work I have examined here are a disparate group.
They range from brilliant intellectuals like Aurobindo, to professional scholars like
B. B. Lai, to what most academics would consider “crackpots,” like P. N. Oak. 1
The primary feature they share is that they have taken it upon themselves to oppose
the theory of Aryan invasions and migrations—hence the label Indigenous Aryanism.
Although I am not fully satisfied with my descriptive label, 1 could find no better
term with which to conveniently refer to my target group. Initially I toyed with the
idea of contextualizing these arguments into a “traditionalist” framework, with the
corollary that such material was an encounter or response to modernity, but not all
the scholars in my target group are traditionally oriented at all; nor do they all by
any means have problems with modernity. “Indian responses to the Aryan migra¬
tion theory” is an obviously inadequate label, since many Indian scholars support
this hypothesis, and a number of Western scholars have begun to contest it. “Dis¬
sident voices from India” failed for similar reasons (the work of several scholars
not resident in India is discussed herein), as did any thoughts of casting the debate
in terms of “Hindu” responses to Western scholarship. Indigenous Aryanism is a
convenient (if somewhat generalized) label, and no more.

A further qualification is in order at this point. The descriptive label Indigenous Aryanism
should, in strict linguistic terms, be Indigenous IndoAryanism, since it specifically refers to
the speakers of the Indie languages. The term Aryan was used to denote the undivided
Indo-Europeans during most of the nineteenth century, for our purposes, the Indo-Aryans
are primarily the Vedic- and pre-Vedic-speaking members of the family. However, Aryan
is often used by the Indian scholars in my survey to denote the Vedic-speaking peoples.
Since the term is, after all, Sanskrit (and Iranian), I have adopted this denotation, in the
context of my descriptive label Indigenous Aryanism. Elsewhere, I use Aryan to refer to all
the Indo-European speakers only in those contexts where the term was used in this gen-

Introduction 5

eral way (particularly when quoting nineteenth-century scholars), and the more precise
term lndo-Aryans to refer to the speakers of Vedic (and related dialects).

My topic, then, is a debate. It is a debate about which most scholars in the West
were unaware until very recently, which in itself says something about the balance of
intellectual power in the academic field of Indology. In order to contextualize the re¬
sponses of what I will call the Indigenous Aryan school—to summarize exactly what it
has been reacting to—the first chapter lays out some of the more prominent features of
the two-hundred-year history of the Indo-European homeland quest in Europe, particu¬
larly as it related to India. The various religious and political exigencies that influenced
much of the scholarship during this period are touched upon, as are the many view¬
points regarding the Indo-European homeland. This should set the scene for the re¬
sponses of my target group—the Indigenous Aryanists—who were observing this intel¬
lectual melee from outside mainstream Western academic circles.

Chapter 2 touches briefly on a variety of discourses that appropriated the Aryan theme
in India in the hope of exploiting it for political or other mileage. The reaction to the
theory by religious intellectuals is also addressed in this chapter, thus providing a brief
Indian parallel to the nineteenth-century political and religious concerns of Europe in
the first chapter. Chapter 3 initiates the analysis of the actual data concerning Indo-
Aryan origins. By the mid-nineteenth century, one of the few tilings regarding the home¬
land that western Indo-European scholars did agree on was that it could not have been
India; wherever the original homeland might have been the lndo-Aryans at least must
have come to the subcontinent from outside. While not the slightest bit concerned with
the homeland obsession of European scholars in general, Indigenous Aryanists soon
reacted to the corollary of the problem when it impinged on the origins of their own
culture. It seemed unacceptable to consider that such an enormously speculative gigantean
and seemingly inconclusive European undertaking should be entitled to make authori¬
tative pronouncements on the early history of the Indian subcontinent. The first voices
of opposition that attempted to utilize critical scholarship to counter the claim that the
forefathers of the Vedic Indians hailed from outside the subcontinent are introduced in
this chapter. The initial objections raised concerned the philological evidence that had
been brought forward as decisive by Western philologist. Since philology was a disci¬
pline that resonated with their own traditional Sntti-epistemologies, and since it focused
on texts in their own ancient language, Vedic Sanskrit, the philological evidence was
the most easily accessible to Indigenous Aryan scrutiny. Moreover, these texts that were
suddenly of such interest to Western scholars happened to be their sacred ones, and
this fueled their concern.

Chapter 4 traces the dethronement of Sanskrit from its initial position as the origi¬
nal protolanguage of all the Indo-Europeans in the opinion of the early linguists, to its
ongoing diminishing status as a secondary language containing a number of linguistic
features that are considered to be more recent than other Indo-European cognate lan¬
guages. Chapter 5 analyzes the evidence for a non-Indo-Aryan linguistic substratum in
Sanskrit texts, which has remained perhaps the principal and, to my mind, most per¬
suasive reason brought forward in support of the Aryan invasions and migrations. The
issue here is: Do the Vedic texts preserve linguistic evidence of languages preceding the
Indo-Aryan presence on the Indian subcontinent? This is an essential aspect of this
debate but one that has been mostly ignored by Indigenous Aryanists. Chapter 6 exam-

6 Introduction

ines various points of view based on the method of linguistic paleontology—one of the
most exploited disciplines used in the homeland quest, and one also fundamental in
insisting that the Indo-Aryans had an origin external to the Indian subcontinent. Here
we will find Indian scholars reconfiguring the same logic and method to arrive at very
different conclusions from those of their Western counterparts. Chapter 7 deals with
the linguistic evidence from outside of India, particularly loan words from the Finno-
Ugric languages, as well as the Mitanni and Avestan evidence, all of which have a direct
bearing on the problem. Here, too, Indigenists have their own way of accounting for
this evidence. Chapter 8 deals with other linguistic issues often utilized in the home¬
land quest and Indo-Aryan origins, such as dialect geography and the implications of
the subgroupings of the various cognate languages. It must be stated immediately that
there is an unavoidable corollary of an Indigenist position. If the Indo-Aryan languages
did not come from outside South Asia, this necessarily entails that India was the original
homeland of all the other Indo-European languages. Indo-Aryan was preceded by Indo-
Iranian, which was preceded, in turn, by Indo-European; so if Indo-Aryan was indig¬
enous to India, its predecessors must have been also. Hence, if proto-Indo-European
was indigenous to India, all the other cognate languages must have emigrated from there.
Chapters six to eight discuss the possibility and problems of a South Asian homeland.

Chapter 9 deals with the relationship between the Indus Valley Civilization and the
Indo-Aryans—a topic that has received a tremendous amount of attention from Indian
archaeologists and historians. The issue to be discussed in this chapter is whether the
Indo-Aryans preceded, succeeded, or coexisted with the inhabitants of the Indus Valley
cities. Chapter 10 outlines some of the scholarship that has attempted to trace the trans-
Asiatic exodus of the Indo-Aryans on their proposed route to India, across central Asia
and chapter 11 examines the problems associated with identifying them in the archaeo¬
logical record within the subcontinent. Chapter 12 examines the various attempts made
to date Sanskrit texts upon which, as I shall argue, a tremendous amount hinges. How
far back can we go with an Indo-Aryan presence on the subcontinent? The final chapter
discusses some of the more modern ideological underpinnings of this debate in India
as different forces compete over the construction of national identity. Other concerns
motivating some of the participants on both sides of the Indigenous Aryan debate will
also be considered in this chapter.

I have left this chapter on ideology until last, in order to present the intervening
chapters on the evidence primarily in the terms, and through the logic and perspectives,
of the various points of view. However, historical data do not tell their own story: they
are interpreted. And interpretation emanates from human cognition that is structured
by each individual’s cultural, religious, political, economic and social circumstances and
choices. People have a reason to contest or reinterpret history. The present volatile situa¬
tion in India has made Western, and many Indian, scholars particularly concerned about
the repercussions of communal interpretations of history. However, although the pro¬
motion of Indigenous Aryanism is undoubtedly extremely important to notions of identity
and to the politics of legitimacy among certain Hindu nationalists, such concerns are
not representative of all the scholars who have supported this point of view. Unfortu¬
nately, the whole Indigenous Aryan position is often simplistically stereotyped, and
conveniently demonized, both in India and in the West, as a discourse exclusively
determined by such agendas. This bypasses other concerns also motivating such

Introduction 7

reconsideration of history: the desire of many Indian scholars to reclaim control over
the reconstruction of the religious and cultural history of their country from the legacy
of imperial and colonial scholarship. In chapter 13 I discuss the manifold concerns
that I perceive as motivating Indigenous Aryanists to undertake a reconsideration of
this issue. I argue that although there are doubtlessly nationalistic and, in some quar¬
ters, communal agendas lurking behind some of this scholarship, a principal feature is
anticolonial/imperial.

On a personal note, l am accordingly sympathetic to the Indigenous Aryan “school”
(if, simply for ease of reference, I might be permitted to reify my motley group as a
school) when I view it as a manifestation of a postcolonial rejection of European intel¬
lectual hegemony (since most of the voices are from India), especially since my analysis
has led me to realize exactly how malleable much of the evidence involved actually is.
This does not mean that the Indigenous Aryan position is historically probable. The
available evidence by no means denies the normative view—that of external Aryan origins
and, if anything, favors it. But this view has had more than its fair share of airing over
the last two centuries, and the Indigenous Aryan position has been generally ignored or
marginalized. What it does mean, in my view, is that Indigenous Aryanism must be
allowed a legitimate and even valuable place in discussions of Indo-Aryan origins.

I am emphatically not sympathetic to the elements of the Indigenous Aryan school
that I perceive as utilizing this debate to construct illusory notions of an indigenous
Aryan pedigree so as to thereby promote the supposed Hindu descendants of these Indo-
Aryans as the original and rightful “sons of the soil” in a modern Hindu nation-state.
As an aside, this is illusory not only from a historico-philological perspective but also
from the perspective of almost the entirety of the philosophical systems associated with
what is known as Hinduism. Vedantic discourse, for one, would consider nationalism
(whether Hindu, American, English, or anything else) to be simply another upadhi, or
false designation, imposed on the dtman out of ignorance (“Hindu nationalism” from
this perspective, is something of an oxymoron). Needless to say, any prioritization of
the Hindus can only be at the expense of the “Other,” namely, the non-Hindu commu¬
nities—specifically Muslims and Christians. Since my task is to be receptive to all ratio¬
nal points of view, including the more cogent interpretations of the Indigenous Aryan
school, there have been many moments when 1 have regretted undertaking this research
for fear that it might be misconstrued and adapted to suit ideological agendas. This
concern very much remains as a dark cloud hovering over what has otherwise been an
intriguing and intellectually very fulfilling research project.

On the other hand, and again on a personal note, I am also concerned at what I
perceive to be a type of Indological McCarthyism creeping into areas of Western, as
well as certain Indian, academic circles, whereby, as will be discussed in chapter 13,
anyone reconsidering the status quo of Indo-Aryan origins is instantly and a priori dubbed
a nationalist, a communalist, or, even worse, a Nazi. Since I have observed that many
scholars, when confronted with “Indigenous” voices of dissension, immediately assume
that it must be just another manifestation of Hindu nationalist discourse (even without
being aware of the linguistic and archaeological issues at stake), a few words on this
issue might be in order at this point.

There is a major difference in focus between nationalism and anti-imperialism al¬
though they overlap in a number of ways. Nationalism involves attempts to concoct

8 Introduction

notions of shared identity in order to unify a variety of individuals and social groups
into a cohesive political and territorial body in contradistinction to other such bodies—
Anderson’s “imagined” communities. Communalism, as understood in the context of
academic discourse about India, also involves attempts to construct political unity be¬
tween individuals and groups, but it is predicated on notions of shared religious affili¬
ation that is distinct from that of other groups who are perceived as identifying with
different religious communities. Anti-imperialism and anticolonialism, on the other hand,
are the opposition, by a group of people, to alien power—often advantaged because of
superior technology—to which it has been subjected against its will. It is a struggle against
an oppressive, and generally stifling, force.

The alien power opposed by the anti-imperialist voices of the Indigenous Aryan school
is intellectual: it consists of the construction of early Indian history by Western scholars
using their “superior technology” in the form of linguistics, archaeology, anthropology,
philology, and so forth. The version of historical events arrived at by these means was
then imposed on the native population in hegemonic fashion. Indigenous Aryanism,
from this perspective, is an attempt to adopt this same technology to challenge the co¬
lonial power (or the heritage it has left behind), to test its foundation, to see how accu¬
rate the Aryan migration hypothesis actually is by examining it with die same equip¬
ment, the same disciplines of archaeology, and so on, that were used to construct it in
the first place. Obviously, in so doing it has been co-opted into a Western critical para¬
digm and has adopted the vocabulary and conceptual structures of the discourse (indeed,
all of the arguments considered here are from English medium publications). 2 It is none¬
theless an attempt to reclaim control over indigenous affairs—in this case die writing of
Indian history—from the power of scholars who are perceived as being motivated and
untrustworthy in their scholarship due to their status as imperialists, colonizers and
rulers (and from the heirs of such scholars). In Gramscian terms, colonial intellectuals
are seen as the “functionaries,” the “deputies,” of the coercive dominant power.

Naturally, perceiving a power to be alien necessitates defining oneself, and others
like oneself, as native, and so anti-imperialism and nationalism often go hand in hand.
Therefore, rejection of the Aryan migration—Indigenous Aryanism—has been co-opted
into both anti-imperial and nationalist discourses. But there is a difference in focus
between these two agendas, and it is this that I wish to emphasize. The focus of those
who promote Indigenous Aryanism for nationalist purposes is to try to show that the
Hindus are an integral, indigenous people who have always had the ingredients required
for a nation-state in the form of a shared culmral past with clear links to the present: the
“Vedic” culture. New nations paradoxically claim to be the opposite of novel, namely,
rooted in the remotest antiquity; nationalists coopt archaeology, philology, and linguis¬
tics in efforts to “prove” that the imagined nation has always existed. In the postcolonial
climate of South Asia, this discourse shares many permeable borders with, and can clearly
spill over into, communal discourse. The focus, in this latter case, is to set up a juxta¬
position between those linked religiously (in however generalized ways) with the Vedic
rubric in opposition to those not affiliated with any notion of Vedic identity—specifi¬
cally the Muslim and Christian minorities. (Here, too, a paradox can arise when disci¬
plines that are empirical in nature are adopted in attempts to prove scripmral authentic¬
ity, which is transcendental in nature, to legitimize notions of identity that are based
upon them.)

Introduction 9

The anti-imperialist concern, in contrast, is the rejection of the Aryan invasion hy¬
pothesis on the grounds that it is an alien intellectual import, assembled by Europeans
as a result of exigencies, initially religious and then imperial, that were prevalent in
nineteenth-century Europe. The theory was exported to the colonies, where it was intro¬
duced by an imperial, colonial power in order to serve imperial, colonial interests. Some
Indigenous Aryanists construe this process as being a conscious one: planned and con¬
spiratorial; others regard it as unconscious and the result of the inevitable bias and self-
centered modes of interpretation that are inherent in the human psyche. The point I
stress in chapter 13 is that not all Indigenous Aryanists are necessarily interested in the
construction of notions of Hindu Aryan greatness or, with some exceptions, in the
promotion of communal agendas. In much of the literature I have read, and in count¬
less hours of interviews, an overwhelming concern of Indigenous Aryanists is to reex¬
amine what is suspected of being a false account of Indian history concocted by Euro¬
pean imperialists—an account that does not correspond to the “facts” even when analyzed
by the modern processes of critical scholarship. In short, the Aryan invasion hypothesis
is seen by many Indian historians as an Orientalist production. As a result, Indigenous
Aryanism can be partly situated within the parameters of postcolonial studies.

Having said all this, I do not intend to suggest that the Indigenous Aryan school is
somehow angelically engaged in the disinterested quest for pure knowledge. There is
no disinterested quest for knowledge. Many Indigenous Aryanists are, indeed, engaged
in the search for self-definition in the modern context. Some are Hindu nationalists,
and some do engage in communal polemics. But much has been written on Hindu
“revisionism” from this perspective; rather than a priori pigeonholing the Indigenous
Aryan school into simplistic and conveniently demonized “communal,” “revisionist,”
or “nationalist” molds that can then be justifiably ignored, this study is an attempt to
analyze and articulate some of the actual empirical objections being raised to tire colo¬
nial construction of Indian pre- and protohistory. This book, accordingly, is primarily
an examination of the empirical, historical evidence—philological, archaeological, lin¬
guistic, and so on—and how this has been interpreted both to support the theory of
Aryan migrations and to contest it. However, since interpretations take place only in a
specific context, a secondary aim is to touch upon (and no more) some aspects of the
religious and political forces, in both Europe and India, that influenced and continue
to influence, the prioritization of certain interpretive possibilities and the exclusion of
others.

A note on method. My intention herein has been not so much to take sides in the
actual debate but to present the interpretations of the evidence from all rational per¬
spectives and point out the various assumptions underlying them—this book is intended
to be a reasonably thorough exposition of the entire problem of Indo-Aryan origins.
Each chapter outlines some of the main features of the history of the data covered in
that particular chapter. Since the relevant matieral is usually so voluminous, however, I
have limited my selection to data that have either attracted responses from the Indig¬
enous Aryan school or that are indispensable to a discussion on the origins of the Indo-
Aryans. This book is not a comprehensive history of the greater Indo-European prob¬
lem, but of the Indo-Aryan side of the family.

I have not hesitated to state my own opinion on the value of some of the arguments
being contested, and my organization and presentation of the material will reveal much

1 0 Introduction

about my own estimation of the merit of some of these points of view. Nonetheless, my
primary project has been to present the debate in its own terms, hence my decision to
quote as much as possible so that the primary voices involved can be heard in their
own right rather than paraphrased. I hold it important that marginalized points of view
that have made a valuable contribution to this issue be brought into a more mainstream
academic context, and it is often edifying to confront as much of the primary tone of the
debate as possible. From the Indigenous Aryan side, the tone of these responses reveals
much about how a historical construct that is taken very much for granted by most of
us—the Aryan invasion/migration theory—is viewed when seen through very different
cultural, religious, and political perspectives.

I have taken material from a wide variety of contexts if I feel that it can contribute to
the debate. I have found that someone, Western or Indian, who can make an astonish-
ingly infantile argument or reveal an alarming lack of critical awareness in one place
can make a penetrating and even brilliant comment in another. Not all the scholars
referred to herein are necessarily schooled in the same intellectual environments, versed
in state-of-the-art academic rhetoric and vocabulary, or familiar with the latest concep¬
tual structures current in Western academia. Nor do all scholars in India have any¬
where near as much access to the latest cutting-edge scholarship or even, sometimes,
basic seminal material as their Western colleagues. I have not rejected any worthwhile
argument even if it is situated in a greater context that many would consider unworthy
of serious academic attention; not all the arguments quoted here are from professional
scholars, but I have allotted space to anyone whom I believe has anything valuable to
contribute to the issue. I have taken it upon myself to wade through a good deal of, to
put it mildly, substandard material in search of nuggets—but nuggets are to be found.

I have also separated my discussion of the evidence from discussion of the religio-
political context of its interpretation: chapters 3 to 12 focus on the former, and chapters
1, 2, and 13 discuss the latter. Talageri the linguist has been critiqued in the chapters
on linguistics, and Talageri the nationalist has been dealt with in the chapter on nation¬
alism. The validity of a particular interpretation of some aspect of the data has not been
minimized because of the author’s overt religious or political bias, however distasteful
they might be to the sensitivities of those of us who do not share those values. Ignoring
a serious attempt to analyze data because of the author’s ideological orientations does
not invalidate the arguments being offered, or make them disappear. On the contrary,
they resurface, often more aggressively because of having been ignored. I have attempted
to analyze, as objectively as possible, any serious interpretation of the data that might
further the task of accurately reconstructing proto-history, while also, in separate chap¬
ters, drawing attention to any ideological agendas that might favor the promotion of a
particular point of view.

I would lie to note that while I have had training in historical Indo-European lin¬
guistics and in South Asia as a linguistic region, I am not an Indo-Europeanist—although
I hold the contributions of this field indispensable to our knowledge of South Asian
pre- and proto-history. I approach this material from the perspective of a historian of
ancient Indian religions and cultures. I beg the indulgence of the specialists and request
that I be forgiven for any errors in technical linguistic detail that they might encounter
in these chapters, and that I be judged, rather, on my more general analysis of the data
and of the conclusions that they can generate in matters pertaining to the origins of the

Introduction

Indo-Aryans. I should also note that while I would like to think that this work might be
of some interest to Indo-Europeanists, if only from a historiographical point of view, or
from the perspective of the history of ideas, it is primarily intended for those involved
in and interested in South Asian studies. With this audience in mind, I have elimi¬
nated all techincal linguistic detail that is not essential to illuminating the general lin¬
guistic principles and theories relevant to the quest for the Indo-Aryans.

In my fieldwork in South Asia, as well as in my research thereafter, I have had ade¬
quate opportunity to discuss the South Asian archaeological evidence with specialists in
the field, although, here again, specialists will likely recognize that I am not an archae¬
ologist. Nor am I a historian of science, despite my lengthy treatment of the astronomi¬
cal evidence. Ultimately, in the hyper-specialized academic culture of our day, it is not
possible for a single individual to have expertise in all of the disciplines and subdisci¬
plines demanded by a topic such as this. I request my critics to consider that no one
can be a specialist in nineteenth-century historiography in Europe, nineteenth-century
historiography in India, Vedic philology, Avestan Studies, historical Indo-European lin¬
guistics, South Asian linguistics, Central Asia as a linguistic area, the archaeology of
Central Asian, the archaeology of the Indian subcontinent, astronomy, and modern
Hindu nationalism, to name only some of the areas covered here. Anyone attempting a
multidisciplinary overview of such a vast amount of material will necessarily need cor¬
rections from the specialists in any of these fields. 1 hope that my efforts have at least
been successful in gathering most of the materials relevant to the origins fo the Indo-
Aryans, shedding some light on why these materials might be contested, and perhaps
invoking further discussions by scholars more qualified then myself in these respective
areas.

Perhaps this is an opportune moment to reveal my own present position on the Indo-
European problem. I am one of a long list of people who do not believe that the avail¬
able data are sufficient to establish anything very conclusive about an Indo-European
homeland, culture, or people. I am comfortable with the assumptions that cognate lan¬
guages evolve from a reasonably standardized protoform (provided this is allowed con¬
siderable dialectal variation) that was spoken during a certain period of human history
and culture in a somewhat condensed geographic area that is probably somewhere in
the historically known Indo-European-speaking area (although I know of no solid grounds
for excluding the possibility that this protolanguage could have originated outside of
this area).

However, regarding homelands, I differ from most Western scholars in that I find
myself hard pressed to absolutely eliminate the possibility that the eastern part of this
region could be one possible candidate among several, albeit not a particularly convinc¬
ing one, provided this area is delimited by Southeast Central Asia, Afghanistan, present-
day Pakistan, and the northwest of the subcontinent (rather than the Indian subconti¬
nent proper). I hasten to stress that it is not that the evidence favors this area as a possible
homeland—on the contrary, there has been almost no convincing evidence brought for¬
ward in support of a homeland this far east. As we shall see, the issue is that problems
arise when one tries to prove that the Indo-Aryans were intrusive into this area from an
outside homeland. In other words, one has almost no grounds to argue for a South
Asian Indo-European homeland from where the other speakers of the Indo-European
language departed, but one can argue that much of the evidence brought forward to

Introduction

document their entrance into the subcontinent is problematic. These are two separate,
but obviously overlapping, issues.

Coupled with the problems that have been raised against all homeland candidates,
these issues have caused me conclude that, in the absence of radically new evidence or
approaches to the presently available evidence, theories on the homeland of the Indo-
European speaking peoples will never be convincingly proven to the satisfaction of even
a majority of scholars. This skepticism especially applies to the theories of some Indian
scholars who have attempted to promote India as a Homeland. I know of no unprob¬
lematic means of re-creating a convincing history of the Indo-Aryan speakers prior to
the earliest proto-historic period, at which time they were very much situated in the
Northwest of the subcontinent (as, of course, were other Indo-European speakers else¬
where). I do not feel compelled to venture any opinions beyond this: how the cognate
languages got to be where they were in prehistory is as unresolved today, in my mind,
as it was two hundred years ago when William Jones announced the Sanskrit language
connection to a surprised Europe. The Indigenous Aryan critique has certainly been
one of the formative influences on my own point of view. In my opinion, this critique
not only merits attention in its own right but, also, perhaps more important, must be
addressed by western scholars, since it is rapidly rising in prominence in the country
whose history is most directly at stake.

Myths of Origin

Europe and the Aryan Homeland Quest

The Indigenous Aryan debate can only be understood in the context of the history of
the greater Indo-European homeland quest in Europe. The purpose of this chapter is to
outline the most prominent features of this history that are most directly connected with
the problems of Indo-Aryan origins. Indigenous Aryanists are almost universally suspi¬
cious of the motives surrounding the manner in which evidence was interpreted and
construed by British and European scholars in the colonial period. It is important to
clearly excavate the various biases that influenced the epistemes of the time before at¬
tempting to consider the evidence itself. This chapter will address some of the more
blatant ideological and religious attitudes of the eighteenth and nineteenth centuries in
the West that co-opted Aryan discourse in some form or fashion. Since there have been
a number of studies focused on the general history of Indo-European Studies, I will
focus only on the aspects of this history that are of particular relevance to the Indian
side of the family.

One common characteristic of Indigenous Aryan discourse is the tendency to dwell
on, and reiterate, the blatant excesses of nineteenth-century scholarship. This is per¬
fectly understandable, and even justified, provided that one proceeds from such analyses
to engage and address the more state-of-the-art views current in our present-day academic
milieus. The function of this chapter is not just to tar and feather all eighteenth- and
nineteenth-century scholars as racists and bigots in order to reject all and any conclu¬
sions formulated in that period as a priori tainted, but to thoroughly acknowledge the
extremities within Western intellectual circles of the time. Only once all that is openly
on the table can one attempt to extrapolate the data from the interpretational constraints
of the time and move on to reexamine it all anew, albeit from within the contextual
constraints of our own. Accordingly, although massive advances were made in the nine-

14 The Quest for the Origin of Vedic Culture

teenth century in the study of ancient India by many dedicated and sincere scholars,
this chapter will focus on the more biased and ideological appropriations of Aryan dis¬
course in Europe. After these elements have been adequately processed and acknowl¬
edged, we can move forward, hopefully somewhat free from the ghosts of the past, to
reexamine the actual evidence from the perspectives of our own present-day postcolonial
academic culture.

Biblical Origins

Scholars and thinkers of the late eighteenth century, enthusiastically pushing forward
the scientific and intellectual frontiers that had become accessible in post-Enlightenment
Europe, found themselves grappling with the historicity of Old Testament chronology.
The discovery, through expanding European colonies, of other cultures claiming pedi¬
grees of vast antiquity; developments in linguistics; and the proliferation of “hard” ar¬
chaeological evidence provoked a drastic reevaluation of biblical narrative in matters of
human origins. Features such as the monogenic descent from Adam, the evolution of
all human language from the monolingual descendants of Noah, and the brief period
that seemed to be allotted to the dispersion of the human race after the Flood became
the subjects of intense debates. As the first pioneering British scholars in India began
to discover Sanskrit texts, the promise of hitherto unknown historical information be¬
coming revealed to Europeans became the cause of both great anticipation and episte¬
mological anxiety.

Sir William Jones, the first Indologist to attempt a serious synchronization of bibli¬
cal and Puranic chronology, exemplifies the tensions of his time. His predecessors, British
scholars John Holwell, Nathaniel Halhed, and Alexander Dow—all associated in vari¬
ous capacities with the British East India Trading Company—had relayed back to an
eager Europe gleanings from Puranic sources that described an immense antiquity for
the human race. 1 These provided the ranks of disaffected Christians, such as the vocif¬
erous Voltaire, with valuable materials with which to attempt to shake off the constraints
of Judeo-Christian chronology and to refute Jewish or Christian claims to exclusive
mediation between man and Providence. Holwell, for one, believed that the Hindu texts
contained a higher revelation than the Christian ones, that they predated the Flood,
and that “the mythology, as well as the cosmogony of the Egyptians, Greeks and Ro¬
mans, were borrowed from the doctrines of the Brahmins” (Marshall 1970, 46). Halhed,
too, seemed to take the vast periods of time assigned to the four yugas quite seriously,
since “human reason . . . can no more reconcile to itself the idea of Patriarchal. . . lon¬
gevity” of a few thousand years for the entire span of the human race (Marshall, 1931,
159). Dow was instrumental in presenting Europe with a deistic image of India whose
primitive truths owed nothing to either Jews or Christians. Such challenges stirred up
considerable controversy in Europe, fueled by intellectuals such as Voltaire adopting
such material in endeavors to undermine biblical historicity.

Naturally, such drastic innovations were bitterly opposed by other segments of the
intelligentsia. For well over a millennium, much of Europe had accepted the Old Tes¬
tament as an infallible testament documenting the history of the human race. Thomas
Maurice, for example, complained bitterly in 1812 about “the daring assumptions of

Myths of Origin

certain skeptical French philosophers with respect to the Age of the World . . . argu¬
ments principally founded on the high assumptions of the Brahmins . . . [which] have
a direct tendency to overturn the Mosaic system, and, with it, Christianity.” Such schol¬
ars were greatly relieved by “the fortunate arrival of. . . the various dissertations, on the
subject, of Sir William Jones” (22-23). Jones was just as concerned about the fact that
“some intelligent and virtuous persons are inclined to doubt the authenticity of the ac¬
counts delivered by Moses.” In his estimation, too, “either the first eleven chapters of
Genesis ... are true, or the whole fabrick of our national religion is false, a conclusion
which none of us, I trust, would wish to be drawn” (Jones 1788, 225).

Eager to settle the matter, Jones undertook the responsibility of unraveling Indian
chronology for the benefit and appeasement of his disconcerted colleagues: “I propose
to lay before you a concise history of Indian chronology extracted from Sanskrit books,
attached to no system, and as much disposed to reject Mosaick history, if it be proved
erroneous, as to believe it, if it be confirmed by sound reason from indubitable evi¬
dence” (Jones 1790a, 111). Despite such assurances, Jones’s own predispositions on
this matter were revealed in several earlier written statements: “I. . . am obliged of course
to believe the sanctity of the venerable books [of Genesis]” (1788, 225); Jones (1790)
concluded his researches by claiming to have “traced the foundation of the Indian empire
above three thousand eight hundred years from now” (145), that is to say, safely within
the confines of Bishop Usher’s creation date of 4004 b.c.e. and, more important, within
the parameters of the Great Flood, which Jones considered to have occurred in 2350

B.C.E.

Such undertakings afford us a glimpse of some of the tensions that many European
scholars were facing in their encounter with India at the end of the eighteenth century;
the influence of the times clearly weighed heavily. However, Jones’s compromise with
the biblical narrative did make the new Orientalism safe for Anglicans: “Jones in effect
showed that Sanskrit literature was not an enemy but an ally of the Bible, supplying
independent corroboration of the Bible’s version of history” (Trautmann, 1997, 74).
Jones’s chronological researches did manage to calm the waters somewhat and “effec¬
tively guaranteed that the new admiration for Hinduism would reinforce Christianity
and would not work for its overthrow” (74). Trautmann notes that, for the most part,
up until the early part of the nineteenth century, British Indomania was excited about
the discovery of Hinduism for several reasons: it provided independent confirmation of
the Bible; its religion contained the primitive truth of natural religion still in practice, a
unitary truth from which the forms of paganism of Rome and Greece were perverted
offshoots; and its arts and cultures were connected to Egypt’s (64).

Jones’s much more lasting contribution, and one generally recognized by linguists as
the birth of historical linguistics, was his landmark address to the Royal Asiatic Society
of Bengal in 1786. This, by constant quotation, has by now become the mangaldcara of
comparative philology:

The Sanskrit language, whatever may be its antiquity, is of a wonderful structure; more
perfect than the Greek, more copious than the Latin, and more exquisitely refined than
either, yet bearing to both of them a stronger affinity, both in the roots of verbs and in
the forms of grammar, than could possibly have been produced by accident; so strong,
indeed, that no philologer could examine them all three, without believing them to have
sprang from some common source which, perhaps, no longer exists: there is a similar

1 6 The Quest for the Origin of Vedic Culture

reason, though not quite so forcible, for supposing that both the Gothick and the Celtick,
though blended with a very different idiom, had the same origin with the Sanskrit; and
the old Persian might be added to the same family, (Jones 1788, 415-431)

Significantly, this statement was almost a paraphrase of a not so well known declaration
made over a century previously (in 1668) by one Andreas Jager in Wittenberg, before
the discovery of the Sanskrit language:

An ancient language, once spoken in the distant past in the area of the Caucasus moun¬
tains and spreading by waves of migration throughout Europe and Asia, had itself ceased
to be spoken and had left no linguistic monuments behind, but had as a “mother” gen¬
erated a host of “daughter languages”. . . . (Descendants of the ancestral language in¬
clude Persian, Greek, Italic, . . , the Slavonic languages, Celtic, and finally Gothic. (Quoted
in Metcalf 1974, 233)

Attempts to demonstrate that the disparate languages of the world stemmed from a com¬
mon source long predated the discovery of Sanskrit. As early as 1610, J. j. Scaliger was
able to distinguish eleven European language groups, such as Germanic, Slavic, and
Romance, and was noteworthy for his time in challenging the idea that these languages
derived from Hebrew—the opinion prevalent in his day. 2

The idea of a common source—initially considered to be Hebrew—for all languages,
which, it is important to note, was always associated with a common people, was taken
for granted by most scholars in Europe until well after the Enlightenment. The idea was
inbedded in the biblical version of history, in which Noah’s three sons, Japheth, Shem,
and Ham, were generally accepted as being the progenitors of the whole of humanity. 3
Prior to the construction of the city of Babel, there was one human race speaking one
language. These linguistically unified and racially integral people were subsequently
dispersed and scattered over the face of the earth. This theme, even when stripped of its
biblical trappings, was to remain thoroughly imprinted in European consciousness until
well into the twentieth century.

In 1768, even before the affinities of Sanskrit with the Indo-European languages had
been officially broadcast by Jones, Pere Coeurdoux foreshadowed much present-day
opinion regarding the point of origin of this language by stating that “the Samskroutam
language is that of the ancient Brahmes; they came to India . . . from Caucasia. Of the
sons of Japhet, some spoke Sarhskroutam” (quoted in Trautmann 1997, 54). Once San¬
skrit had become accessible to British scholars, its connection with the classical lan¬
guages of Europe was suspected even before Jones’s proclamation. Halhed had noted
the possibility a few years earlier. James Parsons, too, physician and fellow of the Royal
Society and of the Society of Antiquities, had also associated Indie with die European
languages in 17 76. In fact, almost two centuries earlier still, the Italian Jesuit Filippo Sassetti,
who lived in Goa in the 1580s, had noted that in the language “there are many of our
terms, particularly the numbers 6, 7, 8 and 9, God, snakes and a number of other things”
(Marcucci 1855, 445). Jones’s status and reputation, however, ensured diat news of this
language connection reverberated through die academic halls of Europe.

Once die discovery of Sanskrit as a language related to the European languages had
been made public, it precipitated post-Enlightenment trends toward disaffiliation from
Genesis. It was a traumatic time for Europe. As Max Muller , who “look[ed] upon the
account of Creation as given in Genesis as simply historical” (1902, 481), later remi-

Myths of Origin l 7

nisced: “All one’s ideas of Adam and Eve, and the Paradise, and the tower of Babel,
and Shem, Ham, and Japhet, with Homer and Aeneas and Virgil too, seemed to be
whirling round and round, till at last one picked up the fragments and tried to build a
new world, and to live with a new historical consciousness” (Muller 1883, 29). This
“new world,” however, retained much of the old, and the biblical framework of one
language, one race was transmitted completely intact. Even after developments in lin¬
guistics had irremediably established the existence of numerous completely distinct lan¬
guage families, and the times no longer required scholars to orient their positions around
a refutation or defense of Old Testament narrative, the biblical heritage continued to
survive in a modified form: the idea of one language family for the superior civilizations
of Europe, Persia, and India—the Aryan, or Indo-European, language family—continued
to be associated with the fountainhead of a distinct people that had originated in a specific
geographical homeland.

The correlation of race and language, an assumption that still occasionally continues
to haunt discussions on the Indo-Europeans, was reinforced by the very vocabulary
adopted by the fledgling science of linguistics: Jager, as we have noted, referred to “mother”
languages generating “daughters.” This genealogically derived vocabulary later became es¬
tablished as standard linguistic parlance by Schleicher, whose basic paradigm of the fam¬
ily tree of languages is still in use, albeit usually in modified form. As Trautmann notes:
“This tree paradigm remains very much the foundation of historical linguistics to this day,
although a kind of willful collective amnesia has tended to suppress its biblical origins.
... In the self-conception of linguistics there came to be a strong tendency to imagine
that its central conceptual structure comes from comparative anatomy and to forget that
it comes from the Bible” (1997, 57). The influence of the Bible, initially overtly and
subsequently in a more inadvertent or subconscious fashion, pervaded the entire field
of Indo-European studies in its formative stages throughout the nineteenth century:

The authors of the nineteenth century were hostages, as we are no doubt too, to the
questions they set themselves. Though they cast aside the old theological questions, they
remained attached to the notion of a providential history. Although they borrowed the
techniques of positivist scholarship, took inspiration from methods perfected by natural
scientists, and adopted the new perspective of comparative studies, they continued to be
influenced by the biblical presuppositions that defined the ultimate meaning of their work.
Despite differences in outlook, Renan, Max Muller, Pictet and many others joined ro¬
manticism with positivism in an effort to preserve a common allegiance to the doctrines
of Providence. (Olinder 1992, 20)

Another instant by-product of the discovery of Sanskrit was that a dramatic new
ingredient had been added to Europe’s quest for linguistic and racial origins. Up to this
point, many European scholars, such as James Parsons in 1767, tended to be “persuaded
that these mountains of Ararat, upon which the ark rested, [were] in Armenia; and that
the plains in their neighborhood were the places where Noah and his family dwelt,
immediately after they left the ark” (10). Even after Sanskrit had been “discovered,”
many scholars would not stray too far from this location. Jones himself was emphatic
that the “primeval events” of the construction of the Tower of Babel and the subse¬
quent scattering of the original monolanguage into different tongues “are described as
having happened between the Oxus and Euphrates, the mountains of the Caucus and
the borders of India, that is within the limits of Iran.” There was no doubt about this,

18 The Quest for the Origin of Vedic Culture

since “the Hebrew narrative [is] more than human in its origin and consequently true
in every substantial part of it.” Therefore, “it is no longer probable only, but absolutely
certain, that the whole race of man proceeded from Iran, whence they migrated at first
in three great colonies [those of Shem, Japhet, and Ham]; and that those three branches
grew from a common stock” (Jones 1792, 486-487).

India, the Cradle of Civilization

Other scholars, however, upon learning of these linguistic (and therefore racial) connec¬
tions of the distant Indie languages, felt that radical alternatives to the Armenian point
of origin had now gained legitimacy. India, in particular, was a popular candidate, espe¬
cially among segments of the intelligentsia in the late eighteenth century and first half
of the nineteenth century, and especially (but not exclusively) on the Continent. As it
had done in classical times, India again captured the imagination of Romantic Europe.
The astronomer Bailly, the first mayor of Paris, was very influential in popularizing Indian
wisdom. In 1777, after some deliberation, he situated the earliest humans on the banks
of the Ganges. Even before Jones’s announcement, Bailly stated that “the Brahmans
are the teachers of Pythagoras, the instructors of Greece and through her of the whole
of Europe” (51). Voltaire voiced his agreement: “In short, Sir, I am convinced that
everything—astronomy, astrology, metempsychosis, etc.—comes to us from the banks of
the Ganges” (Bailly 1777, 4).

The French naturalist and traveler Pierre de Sonnerat (1 782) also believed all knowl¬
edge came from India, which he considered the cradle of the human race. In 1807, the
well-known metaphysician Schelling could wonder “what is Europe really but a sterile
trunk which owes everything to Oriental grafts?” (Poliakov 1971, 11). A year later, the
influential Friedrich von Schlegel argued that “the Northwest of India must be consid¬
ered the central point from which all of these nations had their origin” (505). In 1845,
Eichhoff was adamant that “all Europeans come from the Orient. This truth, which is
confirmed by the evidence of physiology and linguistics, no longer needs special proof’
(12). Even as late as 1855, Lord A. Curzon, the governor-general of India and eventual
chancellor of Oxford, was still convinced that “the race of India branched out and
multiplied into that of the great Indo-European family. . . . The Aryans, at a period as
yet undetermined, advanced towards and invaded the countries to the west and north¬
west of India, [and] conquered the various tribes who occupied the land.” European
civilization, in his view, was initiated by the Indian Aryans: “They must have imposed
their religion, institutions, and language, which later obliterated nearly all the traces of
the former non-Aryan language, or languages, of the conquered tribes” (172-173).
Michelet held that the Vedas “were undoubtedly the first monument of the world” (1864,
26) and that from India emanated “a torrent of light and the flow of reason and Right”
(485). He proclaimed that “the migrations of mankind follow the route of the sun from
East to West along the sun’s course. ... At its starting point, man arose in India, the
birthplace of races and of religions, the womb of the world” (Febvre 1946, 95-96).

According to Poliakov, it was Johann-Gottfried Herder, a Lutheran pastor, who (along
with Kant) placed the homeland in Tibet, who was influential in introducing the pas¬
sion for India into Germanic lands and prompting the imagination of the Romantics to

Myths of Origin

seek affiliation with Mother India. Herder (1803) objected that “the pains that have
been taken, to make of all the people of the earth, according to this genealogy, descen¬
dants of the Hebrew, and half-brothers of the Jews, are contrary not only to chronology
and universal history but to the true point of view of the narrative itself.” As far as he
was concerned, “the central point of the largest quarter of the Globe, the primitive moun¬
tains of Asia, prepared the first abode of the human race” (517-18).

Although it was suspected, in some circles, that the enthusiastic acceptance of India
as the cradle of the human race was a reaction against biblical chronological hegemony,
the position did not initially appear to be without foundation: the new science of his¬
torical linguistics originally seemed to lend some support to this possibility because early
linguists tended to treat Vedic Sanskrit as identical or almost identical to the original
Indo-European mother tongue due to the antiquity of its textual sources. 0ones was
actually exceptionally “modern” in considering Sanskrit, along with the other Indo-
European languages, to be co-descendants of an earlier ancestor language, rather than
the original language.) Linguists of the time believed that Sanskrit showed more struc¬
tural regularity than its cognate languages, which, in keeping with the Romantic worldview,
indicated that it was more “original” than Greek and the other cognate languages. Lord
Monboddo, (1774), for example, felt that he would “be able clearly to prove that Greek
is derived from the Shanscrit” (322). Halhed stated: “I do not ascertain as a fact, that
either Greek or Latin are derived from this language; but I give a few reasons wherein
such a conjecture might be found: and I am sure that it has a better claim to the honour
of a parent than Phoenician or Hebrew (Letter to G. Costard, quoted in Marshall 1970,
10). Schlegel, (1977 [1808]), who played a leading role in stimulating interest in San¬
skrit, especially in Germany, developed the concept of comparative grammar wherein
“the Indian language is older, the others younger and derived from it” (429). Vans
Kennedy (1828) felt the evidence demonstrated that “Sanscrit itself is the primitive lan¬
guage from which Greek, Latin, and the mother of the Teutonic dialects were originally
derived” (196). These ideas were picked up by intellectuals outside the halls of academia:
Blavatsky (1975), the theosophist, claimed that “Old Sanskrit is the origin of all the less
ancient Indo-European languages, as well as of the modern European tongues and dia¬
lects” (115).

Although other languages also provided valuable material, the reconstruction of
the original Indo-European was, in truth, completely dependent on Sanskrit, to which
linguists invariably turned for ultimate confirmation of any historical linguistic for¬
mulation. It seemed logical, at the time, to situate the original homeland in the loca¬
tion that spawned what was then considered to be the original or, at least, the oldest,
language. As Sayce (1883) noted in retrospect, “the old theory rested partly on the
assumption that man’s primeval birthplace was in the East—and that, consequently,
the movement of population must have been from east to west—partly on the belief
that Sanskrit preserved more faithfully than any of its sisters the features of the Aryan
parent speech” (385).

In time, however, the linguist F. Bopp (n.d.) stated: “I do not believe that the Greek,
Latin, and other European languages are to be considered as derived from the Sanskrit.
... 1 feel rather inclined to consider them altogether as subsequent variations of one
original tongue, which, however, the Sanskrit has preserved more perfect than its kin¬
dred dialects” (3). Once the news of this connection seeped out from the ivory towers,

The Quest for the Origin of Vedic Culture

there were clamorous objections raised against the whole linguistic concept of Sanskrit
even being a cognate language (not to speak of the “original” mother language), as will
be discussed later, since the corollary was the outrageous proposal that the people of
Athens and Rome should be considered to have a community of origin with the “niggers”
of India. (Tire kinship of Europeans with Indians was of course, implied by Jones long
before.) But Bopp’s sound scholarship eventually prevailed, the “original tongue” eventu¬
ally became known as Proto-Indo-European, and Sanskrit was demoted to the rank of a
daughter language, albeit “the eldest sister of them all” (Muller 1883, 22).

The term Indo-European was coined in 1816 by the linguist Thomas Young. Rask
toyed with various names such as European, Sarmatic, and Japhetic. Soon, however,
zealous German scholars showed preference for the term Indo-German, popularized by
Julius Klaproth in 1823 (but first used by Conrad Malte-Brun in 1810), on the grounds
that these two languages encapsulated the entire Indo-European-speaking area—the
farthest language to the east being Indie, and to the west, Germanic (Celtic had not
yet been recognized as a distinct language group). This term unsettled the sensitivities
of French and British scholars, who exerted their influence to reestablish the more
politically neutral Indo-European. Bopp preferred to follow their example, since “I do
not see why one should take the Germans as representatives for all the people of our
continent” (quoted in Olinder 1972, 13). The term Aryan was also used extensively
during the nineteenth century and the first part of the twentieth century. This is not
to be confused with Indo-Aryan, which refers exclusively to the Indic-speaking side of
the family. Nowadays Indo-European is the standard term for the whole language fam¬
ily, although some German scholars still prefer Indo-Germanic despite a history of com¬
plaints against it.

As a side note, obviously, there must have been a time prior to any hypothetical
reconstructions of particular protolanguages and protocultures. Accordingly, Proto-Indo-
European is generally defined by most linguists as a language that can be reconstructed
at least dieoretically, at a stage prior to its transformation into distinct languages, and
which reveals a certain cultural environment at an approximate period in human devel¬
opment, in a potentially definable geographic location; linguists acknowledge that this
location can only be identified for that particular (hypothetical) stage of language and
culture. While such generalities are generally accepted, we shall see that anything much
more specific is usually contested (and even these generalities have been challenged). In
any event, in the opinion of most scholars (in the West at least), India soon lost its
privileged position as the point of origin of the Indo-European languages.

A variety of reasons were brought forward to reject the proposal drat India might
have been the original homeland. In 1842, A. W. von Schlegel, in contrast to his brother
Frederik, claimed that “it is completely unlikely that the migrations which had peopled
such a large part of the globe would have begun at its southern extremity and would
have continually directed themselves from there towards the northeast. On the contrary,
everything compels us to believe that the colonies set out in diverging directions from a
central region” (515). He felt that the Caspian Sea area possessed such required central¬
ity. Lassen noted in 1867 that from the countries where the large Indo-Germanic family
resided in ancient times, “India was the most peculiar, . . . and it would be very inexpli¬
cable that no traces of these Indian peculiarities should have been preserved by any
Celtic race in later times, if they had all originally lived in India. . . . Among the names

Myths of Origin

of plants and animals which are common to all these nations there is none which is
native to India” (614). Benfey pointed out that South India was peopled by various
non-Aryan tribes who could hardly have pushed their way through the superior civiliza¬
tion of the Sanskrit-speaking people had the latter been indigenous to the North. These
tribes, therefore, must have been the original natives of India who were subjugated by
the invading Aryans (Muir [I860] 1874, 31 1-312). Muir, summarizing the issues in
1874, fortified all these arguments by arguing that the Sanskrit texts themselves showed
a geographic progression “of the gradual advance of the Aryas from the north-west of
India to the east and south” (xx).

Such arguments were by no means uncontested. In 1841, Mountstuart Elphinstone
objected that “it is opposed to their foreign origin that neither in the code [of Manu)
nor, I believe, in the Vedas, nor in any book ... is there any allusion to a prior resi¬
dence or to a knowledge of more than the name of any country out of India.” Respond¬
ing to some of the arguments that had been brought forward, he argued that “to say
that [the original language] spread from a central point is a gratuitous assumption, and
even contrary to language; for emigration and civilization have not spread in a circle.”
As far as he was concerned, “the question, therefore, is still open. There is no reason
whatever for thinking that the Hindus ever inhabited any country but their present one,
and as little for denying that they may have done so before the earliest trace of their
records or tradition” (97-98).

But, as time went by, such objections soon became far too out of tune with the aca¬
demic consensus, as well as with developing colonial exigencies. Soon after the mid¬
nineteenth century, few scholars were still open to considering either India as the home¬
land of the Indo-Europeans, or protestations regarding the indigenousness of the
Indo-Aryans in the subcontinent. According to Chakrabarti (1976), “it is around the
middle of the nineteenth century that this romantic view of India as sending out roving
bands of ascetics died out. With the Raj firmly established it was the time to begin to
visualize the history and cultural process of India as a series of invasions and foreign
rules” (1967).

The Aryans and Colonial and Missionary Discourse

By the end of the nineteenth century, India was no longer referred to at all as a candi¬
date for the original homeland, and most scholars had situated themselves somewhere
within the parameters of Max Muller’s ([188711985) accommodating opinion that “if
an answer must be given as to the place where our Aryan ancestors dwelt before their
separation, ... I should still say, as I said forty years ago, ‘Somewhere in Asia,’ and no
more” (127). 5 A few peripheral intellectuals noted the change with nostalgia, still hop¬
ing that there would be a reversal in India’s fortunes as a homeland contender. In 1881,
Olcott, a Theosophist, in a lecture given to native audiences in various parts of India,
stated:

The theory that Aryavarta was the cradle of European civilization, the Aryans the progeni¬
tors of western peoples, and their literature the source and spring of all western religions
and philosophies, is comparatively a thing of yesterday. Professor Max Muller and a few
other Sanskritists of our generation have been bringing about this change in western ideas.

22 The Quest for the Origin of Vedic Culture

Let us hope that before many years roll by, we may have out the whole truth about Aryan
civilization, and that your ancestors (and ours) will be honoured according to their deserts.

. . . the Brahmins have their own chronology and no one has the means of proving that
their calculations are exaggerated. . . . We Europeans . . . have a right to more than sus-
pect that India 8,000 years ago sent out a colony of emigrants. (124)

In sharp contrast, the racial scientists, who will be discussed later, recorded the change
of affairs with a note of indignant relief: “In our school days most of us were brought
up to regard Asia as the mother of European people. We were told that an ideal race of
men swarmed forth from the Himalayan highlands disseminating culture right and left
as they spread through the barbarous West.” As far as Ripley was concerned, such philo-
logical ideas represented the dark age of Indo-European studies: “In the days when . . .
there was no science of physical anthropology [and] prehistoric archaeology was not yet
... a new science of philology dazzled the intelligent world . . . and its words were law.
Since 1860 these early inductions have completely broken down in the light of modem
research” (Ripley 1899, 453).

Even during the earlier phase of the homeland quest, when India was still a popular
candidate, many scholars were uncomfortable about moving the Indo-Europeans too
far from their biblical origins somewhere in die Near East. There were those among the
British, in particular, whose colonial sensibilities made them reluctant to acknowledge
any potential cultural indebtedness to the forefathers of the rickshaw pullers of Calcutta,
and who preferred to hang on to the biblical Adam for longer than their European
contemporaries. Even well after Adam was no longer in the picture, there was a very
cool reception in some circles to the “late Prof. Max Muller [who had) blurted forth to
a not over-grateful world the news that we and our revolted sepoys were of the same
human family” (Legge 1902, 710). Again, let us not forget the influence of the times:
many scholars, quite apart from any consideration of India as a possible homeland,
could not even tolerate the newfound language relationship. Muller (1883) again noted
the mood of the day:

They would not have it, they would not believe that there could be any community of
origin between the people of Athens and Rome, and the so-called Niggers of India. The
classical scholars scouted the idea, and 1 still remember the time, when 1 was a student
at Leipzig and begun to study Sanskrit, with what contempt any remarks on Sanskrit or
comparative grammar were treated by my teachers. . . . No one ever was for a time so
completely laughed down as Professor Bopp, when he first published his Comparative
Grammar of Sanskrit, Zend, Greek, Latin and Gothic. All hands were against him. (28)

Unlike some of his contemporaries, Muller was effusive in his admiration for things
Indian (although he never subscribed to an Indian homeland). In his course of lectures
“India: What Can It Teach Us?” (1883), he declared that she was “the country most
richly endowed with all the wealth, power and beauty that nature can bestow,” indeed,
“a very paradise on earth,” a place where “the human mind has most fully developed
some of its choicest gifts, [and] has most deeply pondered on tire greatest problems of
life” (6). Such lavish praise was far too extreme for those who, as Muller himself noted,
would be “horror struck at the idea that the humanity they meet with [in India] . . .
should be able to teach us any lesson” (7).

Muller’s concerns about the reactions that enthusiastic portrayals of India’s superi¬
ority might provoke were not unwarranted. The Indomania of the early British Orientalists

Myths of Origin

“did not die of natural causes; it was killed off’ and replaced by an Indophobia initi¬
ated by Evangelicalism and Utilitarianism, epitomized by Charles Grant and James Mill,
respectively (Trautmann 1997, 99). Well before Muller’s glorifications of India, Grant,
who was very influential in East India Company circles, promoted an aggressive Angli¬
cizing and Christianizing relationship with India, which he provoked by completely
disparaging Indian laws, religion, and character. In contrast to the Orientalists, Grant
([1790] 1970) stressed the absolute difference, in all respects, between the British and the
despicable natives of the subcontinent: “In the worst parts of Europe, there are no doubt
great numbers of men who are sincere, upright, and conscientious. In Bengal, a man of
real veracity and integrity is a great phenomenon” (21). Most significantly, he made
absolutely no reference to the kinship of Sanskrit and the European languages except,
possibly, to note that “the discoveries of science invalidate none of the truths of revela¬
tion” (71). Nor did Grant have any regard for enthusiastic depictions of India. Grant
was quick to criticize scholars who had never even visited India, thereby undermining
the relevance of their scholarship to the real world: “Europeans who, not having resided
in Asia, are acquainted only with a few detached features of the Indian character” (24).

Grant was by no means the first or sole Christian leader to engage in extreme dia¬
tribes against Hinduism—these continued throughout the colonial period. In 1840, the
Reverend Alexander Duff briefly referred to the Aryan commonality by stating that tire
Hindus “can point to little that indicates their high original.” But for the most part he
also simply ranted that they “have no will, no liberty, no conscience of their own. They
are passive instruments, moulded into shape by external influences—mere machines, blindly
stimulated, at the bidding of another, to pursuits the most unworthy of immortal crea¬
tures. In them, reason is in fact laid prostrate. They launch into all the depravities of idol
worship. They look like the sports and derision of the Prince of darkness” (107). In 1882,
William Hastie, principal of the General Assembly of the Church of Scotland’s institu¬
tion in Calcutta, in letters he addressed to “educated Hindus” about their religion, consid¬
ered that “no pen has yet adequately depicted all the hideousness and grossness of the
monstrous system.” Hastie was well aware that Hindu idolatry originated from the same
Aryan stem as that of the Greeks. But the latter had been “recalled from their idolatrous
errors,” while India remained “the most stupendous fortress and citadel of ancient error
and idolatry, . . . paralleled only by the spirits of Pandemonium,” a country whose reli¬
gion consisted of “senseless mummeries, loathsome impurities and bloody barbarous
sacrifices.” It has “consecrated and encouraged every conceivable form of licentiousness,
falsehood, injustice, cruelty, robbery, murder,” and “its sublimest spiritual states have been
but the reflex of physiological conditions in disease” (24-33).

Muller, fully aware of the resentment generated by his Indophilic laudations, took
pains to specify “at once” to the civil servants, officers, missionaries, and merchants
who were actually in the “bazaars” and “courts of justice” of the real-life India that there
were “two very different Indias.” Muller was not unaware of the scathing and disparag¬
ing opinions that his contemporaries in the colonies held regarding the present state of
civilization in India. The India he was referring to “was a thousand, two thousand, it
may be, three thousand years ago” (1883, 7); it was “not on the surface but lay many
centuries beneath it” (1899, 4). The golden age represented a thing very much in the
past. Nonetheless, he did not hesitate in insisting that these ancient Indians “repre¬
sented ... a collateral branch of that family to which we belong by language, that is,

24 The Quest for the Origin of Vedic Culture

by thought, and [that] their historical records . . . have been preserved to us in such
perfect. . . documents . . . that we can learn from them lessons we can learn nowhere
else” (1899, 21).

Such a claim would have been intolerable for the likes of Mill ([1820J 1975), who
had previously censured Jones for indulging in “panegyrics,” finding it “unfortunate
that a mind so . . . devoted to Oriental learning . . . should have adopted the hypoth¬
esis of a high state of civilization in the principal countries of Asia” (109). Mill, too, had
ignored the relationships between the Indian and Western languages, and, like Grant,
insisted on emphasizing the tremendous difference, as opposed to tire Orientalist sense
of kinship, between the British and the Indians. These latter, for Mill, were ignorant
and barbaric and despicable: “No people, how rude and ignorant soever, . . . have ever
drawn a more gross and disgusting picture of the universe” (157). His was a far cry
from the venerable status accorded to the ancient Hindus by Muller and the Orientalists.
The extreme Indophobic discomfort with the connection of Sanskrit with Greek and
Latin was exemplified by the conviction of the Scottish philosopher Dugald Stewart,
who, without knowing a word of die language, proposed that Sanskrit was not a cog¬
nate of Greek, it was Greek. It had been borrowed by the wily Brahmans during
Alexander’s conquest and adopted to keep their conversations inaccessible to the masses
(124). Max Muller (1875), commenting on Stewart’s attempt, again reveals the mood of
the time:

This . . . shows, better than anything else, how violent a shock was given by the discovery
of Sanskrit to prejudices most deeply engrained in the mind of every educated man. The
most absurd arguments found favor for a time, if they could only furnish a loophole by
which to escape the unpleasant conclusion that Greek and Latin were of the same kith
and kin as the language of the black inhabitants of India. (164)

Clearly, the developing pressure to justify the colonial and missionary presence in India
prompted the denigration of Indian civilization, and the shunning of embarrassing cultural
and linguistic ties. Trautmann suggests that such considerations also explain why the
British, despite having primary access to Sanskrit source material, did not pursue the
study of comparative philology. This was to become a predominantly German domain.

The Indo-European language connection, however, was not about to disappear, and
Trautmann masterfully traces the emergence of race science as the resolution of inescap¬
able philological reality with the colonial need for cultural superiority over the natives
of India. It should be noted that up until the middle of the twentieth century, the term
race was used to designate what we would today call an ethnic group (rather than refer¬
ring to the divisions of Caucasian, Negroid, Mongoloid, and so on, as per present usage).
During the nineteenth century, race and nation were more or less interchangeable terms,
but drifted apart in the course of the century as race became more biologized, and na¬
tion politicized. One of the catalysts for the development of race science was that some
of the Orientalists, like Muller, in contrast to the Utilitarians, not only recognized and
appreciated European linguistic and cultural brotherhood with the Hindu Aryans but
also articulated these bonds in terms of racial equality:

No authority could have been strong enough to persuade the Grecian army that their
gods and their hero-ancestors were the same as those of king Porus, or to convince the
English soldier that the same blood was running in his veins as in the veins of the dark Bengalese.

Myths of Origin

And yet there is not an English jury now-a-days, which, after examining the hoary docu¬
ments of language, would reject the claim of a common descent and a legitimate relation¬
ship between Hindu, Greek and Teuton. We challenge the seeming stranger, and whether
he answer with the lips of a Greek, a German, or an Indian, we recognize him as one of
ourselves. Though the . . . physiologist may doubt, ... all must yield before the facts
furnished by language. (Muller 1854a, 29-30; my italics)

The recognition of such racial kinship was repugnant to the ethnologists, who re¬
acted by jettisoning the importance of language and scorning tire Orientalist philolo¬
gists. For them, fairness of skin paralleled highness of civilization, and the Indians were
a beggarly lot who could not possibly be allowed to claim a common racial pedigree,
not to speak of being recognized as “one of’ the British. Isaac Taylor ([1892] 1988)
scathingly exemplifies the antiphilological reaction of the race scientists: “It cannot be
insisted upon too strongly that identity of speech does not imply identity of race, any
more than diversity of speech implies diversity of race” (5-6). Anthropologists such as
himself had scant regard for the Orientalists: “Max Muller, owing to the charm of his
style, his unrivaled power of popular exposition, and to his high authority as a Sanskrit
scholar, has done more than any other writer to popularise this erroneous notion.” Despite
the racist overtones, there was actually a good measure of truth to some of these criti¬
cisms: “Instead of speaking only of a primitive Aryan language, he speaks of an ‘Aryan
race’ and ‘Aryan family.’ . . . more mischievous words have seldom been uttered by a
great scholar” (3-4). As for Muller’s English jury, “the evidence derived from the docu¬
ments of language . . . which might be put before an English jury as to a ‘common
descent’ and a ‘legitimate relationship’ between the negro and the Yankee, would be
more intelligible to the twelve English tradesmen in the box than the obscure evidence
which applies to the case of the Teuton and the Hindu” (6). Taylor attempted to point
out that just as the African American and the European American spoke the same lan¬
guage but were not considered members of the same race, so there were no grounds to
consider the Hindu and the European as being the same race on the basis of the Indo-
European language connection.

Race science was precipitated by the discovery, by the Madras school of Orientalists,
that the South Indian languages were not derivable from Sanskrit. This fact was com¬
bined with certain forced readings of the Vedic texts (which will be discussed in chapter
3) to produce images of the Aryans as white-skinned and dolichocephalic, in contrast to
the dark-skinned, snub-nosed dasas. Despite Taylor’s comments, the one-language-one-
race model inherited from Babel was retained—at least for the white Aryans—and India
was reconstructed as the product of two original races: a fair invading race speaking an
Aryan tongue, and a dark- skinned aboriginal one speaking Dravidian.

For our purposes, both Orientalist philologers and ethnologists agreed on one thing:
Trautmann’s “big bang” of Indian civilization consisted of the impact of Aryan invad¬
ers with the indigenes of India. What the racial theorists succeeded in doing, in their
opposition to the philologists, was to uncouple the common language bond from the
need to identify with the Hindus on any other level whatsoever. The Europeans, as a
race, were now not required to acknowledge any common racial or even cultural bond
with the Hindus: “To speak of‘our Indian bretheren’ is as absurd and false as to claim
relationship with the Negroes of the United States because drey now use an Aryan lan¬
guage” (Sayce 1883, 385). Even the common Indo-European language was presented as

26 The Quest for the Origin of Vedic Culture

being a gift to India from the West, just as it had been to the “negroes of the United
States,” The racial theorists paved the way for the postulate that the Aryans were an
autonomous white race who brought civilization and the Sanskrit language to the differ¬
ent races of India—a development Trautmann holds as pivotal to the political construc¬
tion of Aryan identity developing in Germany. The biblical model of the identity of
language and race still held good for the original white Aryans, but not for the Hindus.
These invading Aryans taught the racially and linguistically distinct natives the Indo-
European language and the arts of civilization. But in so doing, they, in time, lost their
superior status and became racially subsumed by the native population.

There were those in the colonial power who were much more comfortable with these
new developments. The British presence in the subcontinent could now be cast as a
rerun several millennia later of a similar script, but a script that hoped to have a differ¬
ent ending. The British could now present themselves as a second wave of Aryans, again
bringing a superior language and civilization to the racial descendants of the same na¬
tives their forefathers had attempted to elevate so many centuries earlier. Some, drawing
on the findings of racial science, believed that a lesson was to be learned from the ear¬
lier wave of Aryans who had allowed themselves to become degenerate due to their new
environment. Bolstered with the new racial theories, such scholars could now exoner¬
ate themselves for, and indeed insist on, the need for remaining aloof and superior to
their subjects. Thus in Annals of Rural Bengal (1897), W. W. Hunter describes the re¬
tardation of Aryan India, which had become “effeminated by long sloth” because of
miscegenation, “but which could again be regenerated by British Rule” (193). 6 On the
Continent, Gobineau saw India as a warning to Europe of the horrors resulting from
the bastardization of Aryan culture. The trajectory that Aryanism took toward Nazism
in Germany is beyond the scope of the present work, but it has been amply traced in
numerous works, among which Poliakov (1971) is especially noteworthy.

The protestations of the race theoreticians notwithstanding, the equation of language
with race remained entrenched. It certainly did not prevent some British representa¬
tives from exploiting this Aryan commonality in a variety of ways. Henry Sumner Maine
made no bones about the fact “that the government of India by the English has been
rendered appreciably easier by the discoveries which have brought home to the edu¬
cated of both races the common Aryan parentage of Englishman and Hindoo” (18-19).
The headmaster of Marlborough wrote in 1870 that “in coming to Hindostan with our
advanced civilization, we were returning home with splendid gifts, to visit a member of
one common family” (quoted in Maw 1950, 14-15). A few years earlier, one ]. Wilson
insisted that “what has taken place since the commencement of the British Govern¬
ment in India is only a reunion ... of the members of the same great family.” (14-15).
Muller himself (1847) had earlier expressed that “it is curious to see how the descen¬
dants of the same race, to which the first conquerors and masters of India belonged,
return ... to accomplish the glorious work of civilization, which had been left unfin¬
ished by their Arian bretheren” (349).

This reunion, of course, was hardly on equal terms; as Maw (1990) notes, such scholars
“refused to follow the notion to its logical conclusion: that consanguinity entitled con¬
temporary India to a moral parity with Great Britain, and ultimately, to national inde¬
pendence” (36). Far from it. As H. S. Newman (n.d.) was quick to point out: “Once in

Myths of Origin

the end of aeons they meet, and the Aryan of the west rules the Aryan of the east”
(110). Farrer (1870), referring to the “common ancestors from whose loins we both
alike are sprung,” compared the reunion of offspring to that of Esau and Jacob. Accord¬
ing to this association, “from the womb it had been prophesied respecting them that
‘the elder should serve the younger”’ (50; italics in original). Havell (1918) took it upon
himself to speak on behalf of the Indians, who, in his perception, accepted British rule
because “they recognize that the present rulers of India ... are generally animated by
that same love of justice and fair play, the same high principles of conduct and respect
for humanitarian laws which guided the ancient Aryan statesmen and law givers in their
relations with the Indian masses” (vi). Clearly, the Aryan connection could turn out to
be a politically shrewd card to play because “in thus honouring our Aryan forerunners
in India we shall both honour ourselves and make the most direct and effective appeal
to Indian loyalty” (ix).

The Aryan connection proved useful on a variety of occasions and in a variety of
sometimes conflicting ways. Devendraswarup a historian of the colonial period (1993,
36) argues that after the British were shaken by the Great Revolt of 1857, certain indi¬
viduals suddenly found reason to stress their common Aryan bond with the Brahmanas
where others had previously shunned it. Since the Brahmanas were preponderant in
the Bengal Native Infantry, which had taken part in the revolt, there were those among
the British who conveniently began to propagate discourses of Aryan kinship in the
hope of cultivating a sense of identification and allegiance with them (36). Chakrabarti
(1997, 127) notes that the same Risley quoted previously who had voiced such relief
that the new science of racial anthropology exempted the need for Europeans to affiliate
themselves with the Hindu side of the family, did not hesitate in his 1881 Bengal sur¬
vey on the races, religions and languages of India, to allot common Aryan descent lib¬
erally to the Indian groups predominant in the British army such as the Rajputs, Jats,
and Brahmins. The Aryan connection was simply manipulated at will.

Such Aryan commonalty was not only adapted to suit colonial exigencies. Maw (1990)
has thoroughly outlined how certain Christian evangelists also found advantages in dis¬
courses of Aryan kinship. As far as G. Smith was concerned, “the English-speaking Ary¬
ans had been providentially trained to become the rulers of India and evangelizers of India”
since “the youngest civilization in the world was to instruct and correct the oldest” (quoted
in Maw 1950, 35). He noted that “it is not the least of the claims of India on England
that our language is theirs, our civilisation theirs, our aspirations theirs, drat in a very
true and special sense they are our brothers” (35). Samuel Laing held that the “two races
so long separated meet once more. . . . the younger brother has become the stronger,
and takes his place as the head and protector of the family. ... we are here ... on a
sacred mission, to stretch out the right hand of aid to our weaker brother, who once far
outstripped us, but has now fallen behind in the race” (quoted in Maw 1 950, 37).

Along these same lines, a general history' of the subcontinent, written by W. C. Pearce
in 1876, compared the ancient pre-Christian Aryan invasion of the subcontinent to the
modern Christian Aryan one. In his view, the ancient Aryans had descended from the
highlands of Central Asia, bringing with them their language, civilization, and religion,
which far surpassed those of the natives: “IThe Aryan] religion was, in its poetic fan¬
cies, as far exalted above [the native’s] crude systems of worship as the sublime teach-

28 The Quest for the Origin of Vedic Culture

ings of Christianity soar above the doctrines of the code of Menu [sic]” (37). Hastie
(1882) appealed to the “twin branches on the same original Aryan stem”—in this case
the ancient Greek and modern Indie cultures—in order to suggest that the Church could
extinguish the “tenacious survival of the old Aryan world” in modern India just as Paul
had extinguished “the brighter and fairer Hellenism” in the ancient West (25-26).

In contrast to all this, Devendraswarup, (1993), touches on very different Christian
appropriations of the work of the philologists and the discourse of Aryanism: “It seems
that missionary scholars in India had already perceived the potential of the science of
comparative philology in uprooting the hold of the Brahmins” (32). Unlike some of the
discourses noted earlier, other missionaries found it preferable to target the non-Aryan
identity of segments of the Indian populace rather than play up the Aryan commonalty.
The missionaries were having little or no success in converting the Brahmans and up¬
per classes. Devendraswarup finds the scholarly work of missionary intellectuals such
as the Reverend John James Muir and the Reverend John Stevenson readily presenting
the Brahmanas as foreigners who had foisted their Vedic language and texts onto the
aboriginals of India. The idea in this case was to create a sense of alienation from
Brahmanical religion among the lower castes, thereby preparing them for exposure and
conversion to Christianity. Thus Wilson, in a letter to his parents, noted that “the Aryan
tribes in conquering India, urged by the Brahmanas, made war against the Turanian
demon worship. ... It is among the Turanian races,. . . which have no organized priest¬
hood and bewitching literature, that the converts to Christianity are most numerous"
(quoted in Devendraswarup 1993, 35). The Aryan invasion theory proved to be adapt¬
able to a curious mismash of contradictory (but not necessarily competing) interests.

Like some of their administratorial counterparts, still other evangelicals also felt the
need to be very clear about the distinction between the eastern and western branches of
the Aryan family. In 1910, the missionary Slater was quite specific that Christianity had
transcended its Aryan matrix, developing a higher spiritual expression as a result of
influences from the Semitic encounter. India, in contrast, had decayed and remained
“sunk in the grossest superstition.” Slater could not countenace attempts to couple this
sorry state of affairs with western religiosity under a common rubric of Aryan spiritual¬
ity. (Maw 1950, 63). There was no shortage of voices who rejected an Aryanism that
bonded British rulers with their ungrateful subjects (Day 1994, 19).

No one knew what to do about the corollaries of comparative philology. Depending
on their agendas and strategies, British individuals glorified, stressed, minimized,
shunned, or otherwise negotiated in some form or fashion with the Aryan connection.
Scholars went backward and forward, attempting to balance colonial and missionary
exigencies with the academic opinions of the day. For our purposes, whereas India had
been viewed as the homeland of the Aryans and the cradle of civilization at the begin¬
ning of the nineteenth century, by that century’s end, in the opinion of people like
Enrico de Michaelis, it was considered its grave. It seems tempting to suggest that the
concern of many British colonialists during this period was not so much where the
Aryans had come from (there was, after all, no question that England could have been
the homeland), provided they had not come from India, and provided the British did
not need to acknowledge any embarrassing kinship with their Indian subjects. Despite
having primary access to the Sanskrit source material upon which the rest of Europe
was dependent, it was Germany, and to some extent France, but not Britain, that came

M^tfis of Origin 29

to dominate the field of historical linguistics. Sayce was to lament as late as the end of
the nineteenth century that “little is known about it [comparative philology] in England,
for English scholars have but recently awakened to the value and meaning of the work
done by Bopp and Schleicher and Curtius, and have not yet learned that this already
belongs to a past stage in the history of linguistic science” (1883, 385). Although En¬
gland did eventually become a principal center of Sanskrit study for all of Europe (and
had been a pioneer in the early days with people like Jones, Maesden, Leyden, and
Ellis), the British became wary of this new comparative philology. Colonial interests
seem to have superseded this particular pursuit of knowledge.

The dilemma facing the rulers was how to avoid according cultural equality, not to
speak of indebtedness, to the Hindu subjects they intended to govern. The Germans
did not have the same colonial exigencies; on the contrary, philology offered certain
German scholars an opportunity to compensate for their poor showing on the colonial
scene. The British, as expert politicians, were able to turn previously awkward philo¬
logical realities to their political advantage, but it would be well worth exploring the
extent to which they remained wary, in the nascent stages of lndological studies, of the
possibly embarrassing repercussions that might be inherent in exploring the field of
philology. Doubtless other factors were also involved, but philology was nonetheless
very much a German prerogative. 7 There were those among the British who were re¬
lieved to hand the Germans the philological baton of a white Sanskrit-speaking race
that had come into India from somewhere else—anywhere else. And there were those
among the Germans who were happy to take it and run.

Before turning to German Aryanism, it would be unfair to conclude this section
without noting that not all British intellectuals can be generically categorized as arro¬
gant elitists. While it is important to highlight the more extreme versions of Aryan dis¬
course in order to best understand Indian reactions and responses, there were also voices
of moderation and soberness. One cannot tar and feather all nineteenth-century schol¬
ars as racists and bigots. In 1870, Farrer, albeit still convinced of the modern West’s
advancement in civilization, was at least shamed by the excesses of some of his contem¬
poraries in their reactions to the Aryan connection:

Oh! if, instead of calling them and treating them as “niggers”; if, instead of absorbing
with such fatal facility the preposterous notion that they were with few exceptions, an
abject nation of cringing liars, to be despised and kicked, ... if our missionaries had
been tempered sometimes with their religious fanaticism of hatred against idolatry with a
deeper historical knowledge of the religions of the world, . . . then, indeed, the Hindoos
no less than ourselves would have recognized the bond of unity between us because of
the common ancestors from whose loins we both alike are sprung. (48)

Other voices, too, had not hesitated to express disgust at their compatriots:

Is it not something, also, that you all—our Arian friends—should be told, intensely as it
may disgust you, that this Arian Bengali—whom, uncivilly and un-ethnologically, you have
been in the habit of calling a “Nigger,”—is, stubbornly as you may kick against the con¬
viction, your Elder Brother:—one who, much as you may glory in being descended from
certain pig-herding Thegns or piratical Norse Vikings, is, in very truth . . . the represen¬
tative of the pure Arian stock, of which you are a mere offshoot. . . whom it is your duty
to treat with mercy, justice, and forbearance;—as you will have to answer for your dealing
with him to the God and Father of us all. (Blackwell 1856, 548)

30 The Quest for the Origin of Vedic Culture

German Aryanism

Devendraswarup (1993) traces the beginning of Anglo-Germanic collaboration on in¬
tellectual and political planes to the arrival in England of the German scholar Chevalier
Bunsen as ambassador of Prussia in 1841 • The two countries shared an animosity to¬
ward France, which “provided an emotional bond between the rising German nation¬
alism and British imperialism” (32). Bunsen, who was instrumental in bringing Max
Muller to England, made a presentation to the British Association for the Advance¬
ment of Science, promoting the usefulness and importance of the sciences of philology
and ethnology. The Philological Society was established in Britain the very next year,
and the Ethnological Society the year after that. If the British needed prompting from
the Continent, the Germans, in contrast, were very keen to pioneer the new science of
philology. By 1906, the University of Strasbourg could boast a library holding some six
thousand volumes on the Aryans, general ethnology, and related disciplines (Maw 1990,
113). Figueira (1994), in her analysis of Indian thought and the formation of Aryan
ideology, identifies two connected reasons for the initial interest in Vedic scholarship
in Germany: the search for the oldest forms of religion and language, and the inquiry
into the origin and past of the German people through information drawn from an¬
cient Sanskrit sources. These offered a cultural means to restore the ancient greatness of
the Germanic tradition (145-146).

There were very good reasons that some Germans, in particular, took to the rapidly
developing field of philology with such enthusiasm and became the principal promot¬
ers of what Raymond Schwab (1984) has called “the Oriental Renaissance.” The south¬
ern Europeans, all things considered, could point to the grandeur of their ancient Greek
and Latin heritage, and the British could afford to overlook their own potentially em¬
barrassing pedigree problems and bask in the superiority of their colonial and techno¬
logical advances in the modern, real world. It was German national pride that was most
in need of some dramatic infusion from the past. Schwab outlines how, just as the ar¬
rival of Greek manuscripts in Europe after the fall of Constantinople had triggered the
first Renaissance in the fifteenth century, the arrival of Sanskrit texts from India in the
eighteenth and nineteenth centuries produced a “second Renaissance,” with Germans
scholars determined to capitalize on the unique opportunity. After all, if the Germans
could somehow appropriate the mantle of the original Indo-Europeans (which they soon
began to call Indo-Germans), they could then lay claim to being the progenitors of all
subsequent derivative cultures, be they Greek, Latin, or colonial.

Much of Europe, from the eighteenth to the twentieth century, was intensely preoc¬
cupied with racial origins. England, for example, was beset with national identification
problems due to its mixed pedigree of Norman, Anglo-Saxon, and Celtic. Likewise, French
racial theorists had to decide whether a Frankish, Latin, or Gaelic association furthered
their interests, and their affiliation varied according to time and place. In contrast to
such ethnic hybridity, certain members of the German intelligentsia believed they could
lay significant claim to an exclusively indigenous ancestry—claims, they argued, that were
verified by passages in Latin sources as early as Tacitus’s Germania (2.1, 4.1), which
suggested that they were an unmixed and autochthonous people. While some British
historians viewed the multiple invasions and ethnic intermixture that produced British
culture as promoting a “hybrid vigour” that accounted for British preeminence in world

Myths of Origin

affairs, German historians attributed their greatness to the alleged lack of physical or
cultural influences from the inferior people that surrounded them (Trigger 1981, 145).
This latter belief was soon to reappear in various scholarly guises during a phase of
Indo-Germanic studies.

Not only did such individuals believe they were relieved of having to defer to exter¬
nal invaders for their sources of culture or potency, they could actually lay claims to
being the exporters of the most powerful dynasties in Europe: the Swabians of Spain,
Anglo-Saxons of England, Lombards of Italy, Franks of France, and Bavarians of Aus¬
tria were all Germanic tribes. With such credentials, it was a short step for scholars like
Leibniz to claim unabashedly as early as 1690 that “it is certain at any rate that most
inquiries into European origins, customs and antiquities have to do with the Teutonic
language and antiquities” (Leibniz [1690J 1981, 286). Not only this, but since writers of
this period made no distinction between language and race, the Teutonic language, un¬
adulterated by the alien influences that pervaded the other languages of Europe, was, if
not the primeval language of mankind, 8 nonetheless “more natural, . . . more Adamic”
than even Hebrew itself (281). The racial zealousness prominent among this group of
scholars was to have dramatic repercussions in all areas of Indo-European smdies, par¬
ticularly in linguishes and anthropology. The groundwork for postulating a Germanic
homeland for the Indo-Europeans had been laid well before the European discovery of
Sanskrit:

The myth of indigenous barbarian origins developed in Middle Europe, especially in
Germany, which regarded barbarian Europe as the original source of uncorrupted free¬
dom, maintaining individualism and freedom, as opposed to the despotism of Classi¬
cal empires. With the aid of historical linguistics . . . and archaeology ... a national-
historical framework was constructed to legitimate the expanding German nation. Direct
ethnic links were postulated between the prehistoric past and the present on the basis
of ethnic explanations of archaeological cultures. ... It later served as a platform for
racist constructions of a Germanic “Urvolk” to serve the Nazi regime. ... As a conse¬
quence, . . . emphasis on the myth of European oriental origins was toned down.
(Kristiansen 1996, 141)

An oriental origin for the Indo-Europeans was no more compatible with German agen¬
das and aspirations than with British ones.

Before a Germanic homeland could be postulated, however, the consensus regard¬
ing an Asian homeland had to be challenged. Although India had been eliminated as
a potential homeland by the second half of the nineteenth century, scholars still almost
unanimously limited their Aryan debates to other parts of Asia, particularly favoring
Bactria, in present-day Afghanistan, or adjacent areas. Indeed, right up until the end of
tire nineteenth century (by which time other homeland contenders were gaining ground),
scholars such as Monier Williams (1891) still held that “it is probable that one of the
earliest homes (if not the first seat) of the members of the great Aryan family was in the
high land surrounding the sources of the Oxus, to the north of the point connecting
the Hindu Kush with the Himalayas . . . the Pamir Plateau” (4). The first well-known
step toward challenging an Asian homeland in favor of a European one is generally
credited to an Englishman, the ethnologist Robert G. Latham, in 1862. 9

For Latham (1862), “when philologues make the Veda 3000 years old, and deduce
the Latin and its congeners from Asia, they are wrong to, at least, a thousand miles in

32 The Quest for the Origin of Vedic Culture

space, and as many years in time” (620). Latham’s rationale, which survives to the present
day, was that “if historical evidence be wanting, the a priori presumptions must be con¬
sidered. . . . the presumptions are in favour of the smaller class having been deduced
from the larger rather than vice versa ” (611). As a natural scientist, he illustrated this
thesis by comparing language groups to distinct species of reptiles:

Where we have two branches of the same division of speech separated from each other, one of
which is the larger in area and the more diversified by varieties, and the other smaller and
comparatively homogeneous, the presumption is in favour of the latter being derived from the
former, rather than the former from the latter. To deduce the Indo-Europeans of Europe
from the Indo-Europeans of Asia, in ethnology, is like deriving the reptiles of Great Brit¬
ain from those of Ireland in herpetology. (Latham 1851, cxlii; italics in original)

Whatever the value of such reasoning, which will be discussed in chapter 8, Latham
offered a new concept to his fellow scholars—a European homeland for the Aryans.

It was a very small step from zoology and ethnology to physical anthropology, which
was soon pressed into service to identify these original Indo-Europeans in Europe. In
1878, the German philologist L. Geiger was the first to suggest that the Indo-Europeans
were blond, blue-eyed people, and that these traits had become diluted and darkened in
those places where there had been a foreign admixture of genes: “The Indo-Germanic
people remain unadulterated wherever pure blonde traits are best preserved." His logic,
which he bolstered by the same quotes from Tacitus, was that the then available data
showed no evidence of a pre-Indo-European linguistic substratum in north Europe, unlike
other European countries. 10 By the same rationale that India had been eliminated, such
substrata in other areas of Europe suggested that the Indo-Europeans were not native to
these areas but intruders who imposed themselves on preexisting peoples. Continuing
this line of argument, the inhabitants of northern Europe, in contrast to their neigh¬
bors, must have been an indigenous Indo-European race. Since there was no indication
that the Aryans had entered this area from anywhere else, the residents there must have
been the pure descendants of the original Aryans. Their physical traits, by extension,
since they had not been mixed with elements from any other people, must be those of
the original Aryans. The original Indo-Europeans, then, were blond, fair, and blue-eyed.

Dubious interpretations of certain passages in the Vedic texts, which will be exam¬
ined in the next chapters, were then introduced to produce readings of fair, invading
Aryans clashing with snub-nosed indigenous dasas. Armed with such data, another
German, Theodor Poesche (1878), attempted to further the blond cause with an even
more simplistic logic in the same year. He accepted without question that the original
Aryans spoke Indo-European and were blond. Greeks, Italians, and French had the
correct linguistic credentials but were disqualified due to being dark, while some of the
Scandinavians had the right physical qualities but spoke the wrong language (41). The
Germans won by default.

Many in the German nation soon became captivated by the implications of such
possibilities. In 1886, the anatomist and craniologist Rudolf Virchow published a lengthy
report, “The Skin, Hair, and Eye Color of German Schoolchildren” (275-477), based
on a massive investigation involving fifteen million schoolchildren. Questionnaires were
sent out to schools in Germany, Austria, Switzerland, and Belgium to solicit informa¬
tion on the hair and eye color of the students. The statistics showed drat a predomi-

M^tfis of Origin

nance of fair traits occurred in northern Germany and the Scandinavian countries. As
far as German chauvinists were concerned, here was hard scientific proof correlating
the Germans with the pure blond Aryans.

This discussion is circumscribed by focusing primarily on Aryan discourse among
the British and the Germans, since, historically, these have exerted the most influence
on the responses from South Asia, and fuller treatments of the concept of the Aryan
race in Europe during the nineteenth century has been treated in detail elsewhere (Poliakov
1971; Day 1994). But I should at least note in passing that “France has been called ‘the
homeland of racial’ theory” (Day 1994, 1 5). Indeed, Gobineau, whose belief in the blue¬
eyed, blond Aryan did not go down at all well with his contemporaries in his native
France, was soon to have societies named after him spring up all over Germany. In any
event, as scholars from other European countries began to voice their objections to
this reconstructed, blond, Germanic Aryan superman, elements in European anthro¬
pology departments allowed their scholarship to degenerate into a puerile, but fatal,
we’re-more-Aryan-than-you level of discourse. Isaac Taylor ([1892] 1988), while on
the one hand rejoicing that “the whilom tyranny of the Sanskritists is happily overpast”
(332) and that, consequently, philology was no longer the determining method in
Aryan studies, could nonetheless hardly avoid referring to the madness his own dis¬
cipline had unleashed:

The question has been debated with needless acrimony. German scholars . . . have con¬
tended that the physical type of the primitive Aryans was that of die North Germans—a
tall, fair, blue-eyed dolicocephalic race. French writers, on the other hand, have main¬
tained that the primitive Aryans were brachycephalic, and that the true Aryan type is
represented by the Gauls. The Germans claim the primitive Aryans as typical Germans
who Aryanised the French, while the French claim them as typical Frenchmen who
Aryanised the Germans. Both parties maintain that their own ancestors were the pure
noble race of Aryan conquerors, and that their hereditary foes belonged to a conquered
and enslaved race of aboriginal savages, who received the germs of civilisation from their
hereditary superiors. Each party accuses the other of subordinating the results of science
to Chauvinistic sentiment. (226-227)

In 1887, Max Muller ([1887] 1985) joined in the remonstrations against the racial frenzy
enveloping Europe and “declared again and again that if I say Aryas, I mean neither
blood nor bones, nor hair nor skull. . . . How many misunderstandings and how many
controversies are due to what is deduced by arguing from language to blood-relation¬
ship or from blood-relationship to language" (120). Muller may have well felt the need
to stress that “an ethnologist who speaks of an Aryan race, Aryan eyes and hair, and
Aryan blood is as great a sinner as a linguist who speaks of a dolicocephalic dictionary
or a brachycephalic grammar”; after all, it was he who had been a principal cause in
such misconceptions through his earlier remarks on the common blood that the “En¬
glish soldier” shared with the “dark Bengali” (1854a, 29-30). Needless to say, his re¬
traction went largely unnoticed, and the history books recorded the earlier Max Muller
who, for a quarter of a century, had contributed to the idea of a common racial Aryan
ancestry based on a common Aryan tongue. One has only to pick up any book on the
subject from the period to see how effortlessly discourses of language slid into discourses
of race from one sentence to the next: “From a common Proto-Aryan speech we infer
also a common Proto-Aryan homeland. . . . Where was this primitive home from which

34 The Quest for the Origin of Vedic Culture

the Aryan blood went out in so many streams over the earth?” (Widney 1907, 10; my
italics).

Physical anthropology was not the only science invoked to reject the idea of an Asian
homeland. Taylor’s “tyrannical Sanskritists” inspired by the comparative grammars of
pioneering linguists such as Schlegel, Bopp, Grimm, Schleicher, Grassman, Verncr,
Brugmann, and Saussure, developed comparative philology, which, in turn, led to lin¬
guistic paleontology. This methodology was first utilized in a serious way by Adolphe
Pictet in 1859 as one “which utilizes words to culminate in things and ideas” (19). Lin¬
guistic paleontology, which will be discussed in chapter 6, in its basic form, proposes
that cognate words denoting items of material or social culture found in all branches of
a language family, such as ‘wheel’ or ‘horse’, can be used as linguistic evidence to prove
that such items existed in the protoculture of that family. Pictet himself believed his
method pointed to an original homeland in Bactria, present-day Afghanistan. His con¬
temporaries, however, quickly co-opted the new discipline to support the German home¬
land. One of the items that was to be the most amenable in this regard was the com¬
mon beech. Since the beech was well represented in the European side of the family, it
was assumed to have existed in the protolanguage before the various linguistic branches
separated. 12 Since this protolanguage necessitated a protohomeland, scholars such as
Geiger then used this information to draw up maps of the geographic boundaries within
which the beech tree grows—specifically, German-centered Europe—and the Aryan home¬
land was set within this area.

Although linguistics inaugurated the field of Indo-European studies, it did not take
long for archaeology to be summoned to the witness stand to help solve the mystery of
origins (or, in many cases, to provide further “proof’ of predetermined concepts of the
homeland). From abstract philological deductions, the Indo-Europeans had become reified
into a very specific anthropological type living in a very identifiable homeland that ar¬
chaeology, it was hoped, would now physically materialize through the archaeological
record. In 1883, Karl Penka was one of the first scholars to use this method in conjunc¬
tion with linguistics to claim that “the archaeological evidence argues convincingly for a
Scandinavian homeland” based on the Mesolithic culture discovered there (68). More
influential, however, was Gustav Kossina’s defense of a German homeland, in 1902,
which was based on linking the movement of peoples with ceramic changes in the ar¬
chaeological record. Kossina believed the spread of the Corded Ware and Linear Ware
archaeological culmres was indicative of Aryan dispersals. The assumptions upon which
his medrodology was based can be summed up as follows: (1) distinctive artifact types
can be equated with “cultures”; (2) the distribution of such types represents “cultural
provinces”; (3) such provinces can be equated with tribal or ethnic groups; and (4) these
ethnic groups can be identified with historical peoples (provided there has been no major
discontinuity in the archaeological record). Kossina’s assumptions were formative to
Childes’s later work and underpin, to some extent, Gimbutas’s well-known theories.

Once wedded, linguistics and archaeology have not proven to be very comfortable
partners, since their relationship has been marred by acute handicaps in communication.
Archaeology, in the absence of datable inscriptions that are readable, can give no indi¬
cation of the linguistic identity of die members of a particular material culture. Linguis¬
tics, in turn, albeit providing some tantalizing glimpses of material culture through lin¬
guistic paleontology, cannot be easily connected to one specific archaeological entity to

Myths of Origin

the exclusion of others. Moreover, even when an archaeological culture has been con-
vincingly argued to be Indo-European, it has never been accepted uncontroversially as
being the original protoculture.

Archaeology and linguistics have not been the only disciplines involved in the home¬
land quest, and neither, of course, have the Germans been die sole contributors. The
variety of methods used, and the massive differences of opinion expressed over the years,
are truly dizzying. Mallory (1973) has provided an excellent synopsis of some of the
principal homeland hypotheses that have surfaced over the last century and a half. Before
turning to the Indian responses to all of this, it would be useful to glance at a summary
of Mallory’s (1973) outline of some of the other more prominent homeland theories
(prior to the 1970s) that have been articulated by Western scholars. This will give a
clearer picture of the confusion that Indian scholars have had to confront over the de¬
cades and will further help set the stage for their responses.

Two Centuries of Homeland Theories

After the discovery of Sanskrit and the birth of comparative philology, many scholars of
the eighteenth century and the early nineteenth century, as noted previously, maintained
drat India was the original homeland. Rask, however, preferred Asia Minor. Alexander
Murray held Asia to be original, as did Renan, on the basis of the Indo-Iranian literary
material in conjunction with the Bible. John Baldwin opted for Bactria solely on the
basis of the Iranian material. Hehn reconstructed an Indo-European Stone Age pastoralist
and also situated him in Asia, in contradistinction to Benfey, who felt geological evi¬
dence favored Europe as the most ancient abode. Pike, die forerunner of the astronomi¬
cal approach that will be discussed in chapter 9, believed the Rgveda preserved a record
of a vernal equinox that occurred in 5000 b.c.e. in Sogdiana.

Penka elaborated on Poesche’s racial theories, mentioned previously, and bolstered them
with linguistic and archaeological arguments to propose Scandinavia. Charles Morris, also
using racial arguments, envisioned a Proto-Indo-European Mongoloid pastoralist from the
steppe. Isaac Taylor was the first to use the evidence of bodi archaeology and loanwords,
which he believed was indicative of Indo-European origins in Finland. D’Arbois de
Jubainville felt cognate Indo-European words such as house and door indicated a sedentary
life (as opposed to the usually depicted Indo-European pastoral-nomadic one) that flour¬
ished near the Oxus River adjacent to the great Asian civilizations. The anthropologist
Brinton felt the Indo-European languages were the result of the coalescence of a variety of
languages situated in western Europe, T. H. Huxley considered the Indo-European speak¬
ers to be blond dolicocephalics living between the North Sea and the Ural Mountains,
while Otto Schrader, rejecting the racial input, situated the homeland in south Russia
which could accommodate both agriculture and pastoralism. Schmidt foreshadowed some
recent scholars by suggesting that the Indo-Europeans must have been adjacent to Babylon
based on a shared duodecimal numbering system. This idea was opposed by Hirt, who
preferred the Baltic area on die basis of linguistic paleontology.

Ripley’s work, at the turn of the twentieth century, characterized the increasing use
and promotion of archaeological evidence in homeland proposals; he was followed by
Paape, who typified the rising vigor of the German Urheimat (original homeland) school.

36 The Quest for the Origin of Vedic Culture

Whitney disagreed with the latter, arguing that Germany was too cold and forested,
preferring south Russia. This was also the area of choice for Sigmund Feist on the grounds
of linguistic paleontology. Harold Bender differed somewhat and considered Lithuania
more suitable, since Lithuanian was the most conservative Indo-European language, while
P. Giles, who also believed that the Proto-Indo-European were agriculturists and not
nomads, felt that they were more likely to have come from Hungary. Gordon Childe,
however, felt that agriculture was a European innovation and located the homeland in
southwest Russia on archaeological grounds. A. H. Sayce, who was to change his mind
repeatedly, situated the homeland in Asia Minor based on the Hittite evidence in con¬
trast to T. Sulimirski, another forerunner of Gimbutas, who saw the Indo-Europeans as
Russian nomads who buried their dead in burrows called kurgans, and who invaded
Europe—a theory supported by Georges Poisson.

I have discussed the partiality of German scholars to a German homeland in oppo¬
sition to an eastern or steppe one, and such views were further propagated by Walter
Schulz, on the grounds of central European archaeological continuity; by Gustav Neckel,
on the basis of the proto-Indo-European steed being the European horse; by Hans Heger,
on archaeological evidence; and by Fritz Flor, also on theories of horse-riding origins.
Wilhelm Koppers disrupted the Germanic tendencies somewhat by advocating West
Turkestan as a homeland on the basis of the Indo-European connections with the Altaic
peoples, only to be succeeded by Julius Pokorny, who defended a Central Europe home¬
land on the old grounds that Germany showed no evidence of non-Indo-European
substratum influence and on other evidence.

C. Uhlenbeck located the homeland in the Aral-Caspian steppes on grammatical
grounds, while N. S. Trubetzkoy rejected the whole concept of a protolanguage, prefer¬
ring to speculate that the originally different Indo-European languages had developed
similarities through geographic proximity'. Anything approaching a homeland, he be¬
lieved, would be found in an area nearer the Finno-Ugrics, due to the structural ana¬
logues of Indo-European with the Uralic and Caucasian families. Stuart Mann predi¬
cated a north or northeast European homeland on the grounds of comparative folklore,
not far from where Ernst Meyer decided to siniate his nonnomadic, pig-keeping, seden¬
tary Indo-Europeans. Anton Scherer tried to satisfy everybody by proposing a large area
that stretched west from the Urals, right across central and south Russia, and up to the
Baltic, since this was an area that could accommodate both nomadic and sedentary
cultures. Wilhelm Schmidt narrowed this area back down to central Asia on the grounds
of the domesticated horse; Georg Solta, like Trubetzkoy, also rejected the whole concept
of a protolanguage; and Thieme resurrected the German Urheimat position, again on
the basis of linguistic paleontology.

Alfons Nehring analyzed the non-Indo-European influences that he thought supported
an Indo-European origin somewhere between the Altaic people and the Caucasus; Hugh
Hencken, like Scherer, proposed a large compromise zone in southeast Europe and
southwest Russia, while Weriand Merlingen supported Schrader’s thesis, and Gustav
Schwantes returned to the German Urheimat theme. Bosch-Gimpera and G. Devoto
advocated the Danubian cultures of central Europe as the most likely homeland candi¬
dates, to which Wolfgang Schmid concurred, proposing a Baltic Homeland on the
grounds of Hans Krahe’s work on river hydronomy (which suggested that the rivers in
Europe had old Indo-European names).

Mytfis of Origin

The original Aryans have been reconstructed as being nomadic pastoralists, seden¬
tary agriculturists, dolichocephalic, brachycephalic, blond and fair, and brown-haired
and dark. The Indo-European homeland has been located and relocated everywhere
from the North Pole to the South Pole, to China. It has been placed in South India,
central India, North India, Tibet, Bactria, Iran, the Aral Sea, the Caspian Sea, the Black
Sea, Lithuania, the Caucasus, the Urals, the Volga Mountains, south Russia, the steppes
of central Asia, Asia Minor, Anatolia, Scandinavia, Finland, Sweden, the Baltic, west¬
ern Europe, northern Europe, central Europe, and eastern Europe.

Quite apart from the massive divergence of opinion among different scholars, on
occasion even individual scholars cannot make up their own minds. In 1875, A. H.
Sayce could “assume it has been proved that their original home was in Asia, and more
particularly in the high plateau of the Hindu Kush” (389). In 1883, he was “much at¬
tracted by the hypothesis of Poesche which makes the Rokytno marshes the scene” (385).
According to Sayce, “the evidence now shows that the districts in the neighbourhood of
the Baltic were those from which the Aryan languages first radiated . . . though Penka
rejects it with disdain” (385). Four years later Sayce (1887) had changed his mind to
“conceive Dr. Penka to have been . . . right in identifying ‘the Aryans’ with the
dolicocephalic inhabitants of central and North-western Europe . . . and thus remove
the necessity of our falling back on Dr. Poesche’s theory, which traces . . . the white
race to the Rokytno marshes of Russia” (52-53). The same man ended up convinced,
in 1927, that the facts revealed that Asia Minor was the actual homeland (Mallory
1989, 143).

Not everyone has been lured by the quest tor the Indo-Europeans; there has been
a long history in the West of scholars repeatedly voicing criticisms over the years:
“The ‘problem’ is primarily in the head of Indo-Europeanists: It is a problem of in¬
terpretative logic and ideology. We have seen that one primarily places the IE’s in the
north if one is German, ... in the east if one is Russian, and in the middle if, being
Italian or Spanish, one has no chance of competing for the privilege” (Demoule 1 980,
120). Earlier still, in 1948, Hankins articulated the level of disillusionment in his
time:

Skepticism in scholarly circles grew rapidly after 1880. The obvious impossibility of actu¬
ally locating the Aryan homeland; the increasing complexity of the problem with every
addition to our knowledge of prehistoric cultures; the even more remote possibility of
ever learning anything conclusive regarding the traits of the mythical “original Aryans”;
the increasing realization that all the historical peoples were much mixed in blood and
that the role of a particular race in a great miilange of races, though easy to exaggerate, is
impossible to determine, the ridiculous and humiliating spectacle of eminent scholars
subordinating their interests in truth to the inflation of racial and national pride—all these
and many other reasons led scholars to declare either that the Aryan doctrine was a fig¬
ment of the professional imagination or that it was incapable of clarification because the
crucial evidence was lost, apparently forever. (265)

Even Mallory (1989), who has been the most prolific scholar in quest of the Indo-
Europeans, is moved to quip: “One does not ask ‘where is the Indo-European home¬
land?’ but rather ‘where do they put it now?”’ (143). He concludes his summary of the
various Indo-European homeland theories by noting that “the cynical have been tempted
to describe it as the phlogiston of prehistoric research” (1973, 60).

38 The Quest for the Origin of Vedic Culture

Present-Day Homeland Hypotheses
Gimbutas and the Kurgan Theory

Interest in the Indo-European homeland problem seemed to wane during the decades
after the war. Sherratt (1988) wonders whether perhaps many prehistorians avoided
this issue in reaction to the political abuses of archaeology under the Nazis and the
explicit racism that was the ultimate outcome of the Romantic search for ethnic origins
in Germany (459). The last two decades or so, on the other hand, have seen an explo¬
sion of renewed interest. However, even after two hundred years of intense specula¬
tion, there is still no significant consensus regarding “where they put it now”; the
situation has hardly changed. Referring to the panoply of present-day opinions,
Mallory’s (1997) most recent conclusion is: “We have different sub-regions of an early
IE world, scattered in space from the Baltic to Anatolia and east across the European
steppe. ... To unify these disparate geographical elements together into a single ‘uni¬
fied theory’ seems to be as distant to those seeking such a goal in Indo-European
studies as it is for physicists” (11 7).

Two or three current theories will illustrate the extent to which methods and conclu¬
sions vary. The Caucasus area has received considerable attention as a likely homeland,
although this proposal is receiving increasing criticism of late. Marijas Gimbutas, for
well over half a century, has proposed an Uralic/Volgan steppe homeland. This is based
on an archaeological culture labeled the Kurgan culture, which is distinguished by a
specific type of burial mound found in that region (kurgan is the Slavic and Turkic term
for ‘barrow’). Gimbutas argues that this culture can be adequately correlated with Indo-
European culture as revealed by comparative philology. 13 Crucial to Gimbutas’s theory
is the thesis that these Indo-Europeans were mounted warriors with male-associated

Gimbutas’s Indo-European homeland.

My its of Origin

thrusting weapons who, being the first people to domesticate the horse, used their martial
advantages to impose their culture on their neighbors in Old Europe. Gimbutas’s read¬
ing of the archaeological record reveals a dramatic upheaval in the life of the peaceful,
egalitarian, agrarian, matriarchal and artistic Europeans of the fifth and fourth millen¬
nium B.c.e. as a result of these violent Kurgan intrusions.

As Mallory (1989) points out, however, scholars have argued that “almost all of the
arguments for invasion and cultural transformations are far better explained without
reference to Kurgan expansions, and most of the evidence presented so far is either
totally contradicted by other evidence or is the result of gross misinterpretation of the
cultural history of . . . Europe” (185). Refreshingly undogmatic about the whole home¬
land enterprise, he nonetheless holds that her homeland is the least problematic of the
various options. 14 Anthony (1995b) Although also accepting the steppe as the Indo-
European Homeland, claims that Gimbutas’s Old-Europe theories would have passed
unnoticed had they not caught the attention of ecofeminists. He points out that many
Copper Age settlements in Old Europe, Gimbutas’s “gynocentric utopias,” were actu¬
ally heavily fortified, and some of the weapons in the Kurgan graves were probably imports
from “peaceful” Europe. He accuses Gimbutas of taking archaeological items out of their
proper context and finds deforestation and environmental degradation a more likely
culprit for the transformation of Old Europe. Schmitt (1974), too, has pointed out ob¬
jections to Gimbutas’s methods: “Here is the radical error: With the methods of lin¬
guistic paleontology anything may be proved as Proto-Indo-European, but it can not be
proved as typically Proto-Indo-European. Such reconstructions do not exclude the pos¬
sibility that this thing, institution or whatever it may be, may have existed also in other
language families” (283).

Renfrew (1987) is also completely dismissive of the validity of linguistic paleontol¬
ogy, as many scholars have been, since he feels that this method could accommodate
“almost any homeland theory” (86). Through this perspective, since the south Russian
homeland was originally established on the grounds of linguistic paleontology, any at¬
tempt to examine the archaeological evidence in an area defined by a suspect method is
itself a priori suspect. Moreover, he notes (1999) that the earliest evidence of mounted
warriors is not until 1000 b.c.e., far too late for Gimbutas’s theory and hence “without
such military possibilities, the whole explanatory basis for the supposed 'kurgan’ inva¬
sion at the beginning of the Bronze Age disappears” (268). Zimmer (1990b) points out
that while the Proto-Indo-Europeans knew the horse, there is no proof that they neces¬
sarily knew the domesticated horse, and there is no linguistic evidence that they fought
on horseback (316-317). Renfrew (1998) pursues this line of argument by arguing that
anyway the horse was not of military significance in Europe until around 1000 b.c.e.;
during the Iron Age, which “undermine(s) the principal rationale sustaining the ‘Kurgan
Migration’ theory for the origin of the Indo-European languages” (207).

From a completely different angle, Dolgopolsky (1990-93) states that the “loan con¬
nections between IE and Semitic prove that speakers of proto-IE and proto-Semitic lived
in territorial vicinity, which would have been impossible had we accepted Gimbutas’s (and
Mallory’s) hypothesis of the Proto-Caspian steppes as the homeland of proto-IE” (244).
Moreover, he notes that there are no loans from Proto-Indo-European into the Finno-
Ugric languages or vice versa (although there are many from the later Indo-Iranian lan¬
guages), which he believes should have been the case had the Proto-Indo-European been

40 The Quest for the Origin of Vedic Culture

neighbors with these languages in the Pontic-Caspian steppes (245). Renfrew can find no
convincing evidence for the motive behind such a Kurgan spread: he notes that central
and western Europe are not really suited to nomad pastoralism.' 5 He also rejects what he
considers to be a migrationist view—a view, as we have seen, going back to Kossina, that
treats any innovation in the archaeological record, such as a new pottery form or decora¬
tive style, as indicative of a migration of people and a displacement of language. Finally,
in a response to his own critics, he points out that while the language of the Kurgan people
may well have been Indo-European, “there is no logic in the inference that the belief sys¬
tem of the previous ‘Old Europe’ phase need be non-Indo-European simply because its
iconography is different from what follows” (Renfrew 1990, 23).

The most recent critique of Gimbutas at the time of this writing argues that her home¬
land theory is completely incompatible with the linguistic evidence: “Although the hy¬
pothesis claims to have answered the question of the origins of Proto-Indo-European
(PIE), an inherently linguistic construct, there are serious problems with this hypothesis
from a linguist’s point of view” (Kretl 1998, 267). Krell compiles lists of items of flora,
fauna, economy, and technology that archaeology has accounted for in the Kurgan cul¬
ture and compares it with lists of the same categories as reconstructed by traditional
historical-Indo-European linguistics (a method which will be discussed in detail in chapter
6). She finds major discrepancies between the two. 16 She also underlines the fact, which
will be dealt with at length in later chapters, that we cannot presume that the recon¬
structed term for ‘horse’, for example, referred to the domesticated equid in the protoperiod
just because it did in later times. It could originally have referred to a wild equid, a
possibility that would undermine the mainstay of Gimbutas’s arguments that the Kurgan
culture first domesticated the horse and used this new technology to spread surround¬
ing areas, thus spreading the Indo-European languages.

Krell (1998) further points out that the Proto-Indo-European had an agricultural ter¬
minology and not merely a pastoral one; “thus, one can hardly argue, based on the
linguistic data, that Gimbutas’ Kurgan economy is unmistakenly reflected in PIE” (274).
As for technology, “there are also equally plausible reconstructions such as *nau . . .
which suggest knowledge of navigation, a technology quite untypical of Gimbutas’ Kurgan
society” (274). Krell concludes that

Gimbutas seems to first establish a Kurgan hypothesis, based on purely archaeological
observations, and then proceeds to create a picture of the PIE homeland and subsequent
dispersal which fits neatly over her archaeological findings. The problem is that in order
to do this, she has had to be rather selective in her use of linguistic data, as well as in
her interpretation of that data. This is putting the cart before the horse. Such an unsys¬
tematic approach should have given her linguistic proponents real cause for question¬
ing the relevance of her theory, especially if one considers that, by virtue of its nature,
the study of PIE is first and foremost a matter for linguistic, not archaeological investi¬
gation. (279-280)

Renfrew and the West Anatolian Homeland

Rather than an aggressive, mounted seminomad from the steppes, Renfrew (1987) con¬
structs a peaceful, sedentary agriculturist from Anatolia as his Indo-European par excel¬
lence. No two accounts could be metre at odds than Gimbutas’s and Renfrew’s. For the

Myths of Origin 41

Renfrew’s Indo-European homeland with two hypothoses (A & B) for the trajectory of Indo-
Iranian.

latter, the spread of the Indo-European languages was achieved not by Gimbutas’s horse-
riding warriors but by the gradual spread of farming techniques. Moreover, faulting the
circular reasoning that he feels is employed in support of the fifth millennium b.c.e.
date commonly assigned to the united Proto-Indo-Europeans, Renfrew proposes a date
around 7000 b.c.e., based on paleoethnobotanical dates for the introduction of farm¬
ing into Europe from Anatolia. Quite apart from his assignment of a much earlier date
than his peers are comfortable with, his theory has been particularly criticized on lin¬
guistic grounds. Place-names, for example, are the most conservative and durable part
of a language and are generally retained even by other intruding linguistic groups that
might superimpose themselves on an area, yet place-names in Anatolia are unanalyzable
as Indo-European. Crossland (1988) wonders why, if Anatolia were the original home¬
land, the Indo-European language discovered there, Hittite, had so little impact on
neighboring languages and was itself a minority language heavily influenced by die non-
Indo-European languages in the environs such as Hurrian and Hattie. Renfrew has also
been criticized for not accounting for the language connections between Finno-Ugric
and Indo-European, which, if correct, would suggest the neighboring relationships of
these languages in the Volga Valley.

Zimmer (1990a, 319) notes that there are no words reconstructable in Proto-Indo-
European for wheat and barley, Renfrew’s basic agricultural crops. This objection was
elaborated upon by Haarmann (1994), who notes that Proto-Indo-European would have
had a full-fledged agricultural vocabulary in the protolanguage were Renfrew’s theory to
hold good. This is not the case, since the European languages and the Indo-Iranian
ones seem to have developed separate sets of agricultural terms after the dispersal.
Haarman also notes that were Renfrew correct any such hypothetical Proto-Indo-Euro¬
pean agricultural terms would have been borrowed by language families adjacent to

42 The Quest for the Origin of Vedic Culture

Anatolia, such as Semitic, which is not the case. Also, such terms would have surfaced
in Greek, but the agricultural terms in this language are non-Indo-European. Archaeol¬
ogy, too, does not support an intrusion into Greece from Anatolia. 17

Lamberg-Karlovsky (1988, 2) points out that if agriculture was invented by Indo-Eu¬
ropeans, then people in a broad area from the Sinai to the Iranian Plateau must have
spoken Indo-European at 7000 b.c.e. at the very latest. Moreover, the earliest domesti¬
cation of cereals began at least by 9000 b.c.e., not in Renfrew’s 7000 b.c.e. and in the
Natufian culture best known from archaeological work in Israel not in Anatolia. As far
as he is concerned, “the complex interaction of numerous communities . . . spread over
a large area of the Near East forms the background to an understanding of agricultural
origins, ... a process which took several thousand years, involved numerous distinct
archaeological cultures that no doubt spoke a variety of different languages.” He has
little sympathy for such a complex process being “simplified by Professor Renfrew to
the ludicrous formula 7000 b.c.e. Anatolia = farming = Indo-Europeans” (2).

Gamkrelidze and Ivanov and the East Anatolian Homeland

An Anatolian homeland, albeit based on very different methods and located farther
east and much later than Renfrew’s homeland has received the support of the linguists
Thomas V. Gamkrelidze and V. V. Ivanov over the years (1983a, b, 1985, 1990,a b,
1995). They utilize the evidence of loanwords, particularly Semitic ones, to situate their
homeland adjacent to the Middle East—not far from where, well over two hundred years
ago, James Parsons had put the descendants of Noah. Using linguistic paleontology,
these linguists have reconstructed a mountainous landscape for the Indo-European
homeland, based on the many cognate words for high mountains, mountain lakes, and

Gamkrelidze and Ivanov’s Indo-European homeland.

Myths of Origin

rapid rivers flowing from mountain sources. They argue that this is incompatible with
the plains of central Europe but quite suitable for the area around eastern Anatolia backed
by the Caucasus. In addition, Gamkrelidze and Ivanov present a much warmer, and
exotic, southern landscape and climate—replete with monkeys and elephants—than
Gimbutas’s cold, austere, northern scenario.

However, many of the Semitic loanwords that are fundamental to Gamkrelidze and
Ivanov’s situating of the homeland next to the Near East have been challenged by
D’iakonov, who defends a Balkan-Carpathian homeland and finds particular problems
with the details of the tribe migrations in their model. Manczak (1990) points out that
the Indo-European language Armenian, which is spoken more or less in Gamkrelidze
and Ivanov’s putative homeland, shows signs of massive substratum influence, indicat¬
ing that it was not indigenous to the area. Moreover, the archaeologists—such as Gimbutas
and Mallory—feel archaeological evidence has not been sufficiently accounted for in
Gamkrelidze and Ivanov’s model, which is primarily linguistic, and that there is insuf¬
ficient archaeological evidence accompanying the postulated spread of Indo-European
languages from this area. Where Gimbutas relies almost exclusively on archaeology and
demonstrates a panoramic mastery over a mass of archaeological minutiae in her series
of articles, Gamkrelidze and Ivanov awe their readers with an encyclopedic exhibition
of comparative linguistic detail in a thousand-page tome that requires a separate volume
just to incorporate the bibliography and index. Yet these two scholars have reconstructed
dramatically different locations and cultures for the homeland. Still another recent and
innovative theory situating the homeland in Bactria, that of Nichols, will be discussed
at length in later chapters.

Conclusion

Scholars hardly agree on even the most basic details of the Indo-European—any more in
the present than they did in the past. Such lack of scholarly consensus on even basic
points perhaps epitomizes the history and culture of the Indo-European homeland quest
more than most other comparable undertakings. Typically, a convincing and detailed
proposal offered in one field (e.g., archaeology) is undermined by evidence from an¬
other (i.e., linguistics) and vice versa. Any attempt to isolate or highlight one aspect of
the data as paramount is inevitably countered by contradictory conclusions produced
by other factors. A convincing picture has yet to emerge from the totality of evidence
despite significant advances in the relevant disciplines. 18

Even within disciplines, archaeologists such as Renfrew, Gimbutas, and Mallory, just
like their predecessors in the last century or so, significantly disagree with each other de¬
spite their sharing a common field. As for dialogue across disciplines, that is, between
archaeologists and linguists, even when a common language can be agreed upon, basic
points of reconstructed Indo-European culture are often not. Gimbutas’s (1985) recon¬
structed Proto-Indo-European, as noted, is a seminomadic pastoralist: “Neither archaeol¬
ogy nor linguistic evidence supports the hypothesis that the proto-Indo-European culture
was in the stage of developed agriculture” (186). In the same journal, the linguist D’iakonov
(1985) states: “The Proto-Indo-Europeans were not nomads: their well-developed agricul¬
ture and social terminology testifies against this; and so does history” (148).

44 The Quest for the Origin of Vedic Culture

Both linguists and archaeologists sometimes claim primacy for their discipline over
locating the Indo-Europeans. The linguist Rudiger Schmitt (1974) notes that several
prehistorians have assigned a variety of quite different homelands to the Indo-Europeans
that all seem to fit the presently available linguistic data reasonably well. Since the Indo-
Europeans are a linguistic concept, he argues that it is ultimately meaningless to search
for archaeological evidence of their existence until linguistics and philology can narrow
down their whereabouts much more precisely. Only then should archaeology be called
in to identify them materially. Otherwise, “prehistorians have no difficulty in finding
evidence in many locations which will fit the existing linguistic data” (285). Gimbutas
(1974), despite giving some token recognition to the “gold mine” of two centuries’ worth
of linguistic research, launches her response (which deals almost exclusively with ar¬
chaeological material, as do all her publications) with an aggressive claim: “It is quite
obvious that the solution of the PIE origins—on a spatial and temporal basis—is in the
hands of archaeologists” (289). Responding to this remark, in turn, another linguist,
Dolgopolsky, who also locates the homeland in Anatolia on linguistic grounds, retorts:
“I completely disagree with Gimbutas’ statement. ... It is, on the contrary, far from
self-evident how archaeologists utilizing the non-linguistic means at their disposal can
determine what language the bearers of some Pit-Grave culture or Battle-Axe culture spoke”
(1987, 7). He continues to give rein to his impatience: “It is here maintained that the
linguists are ultimately responsible for determining the geographical and cultural pa¬
rameters of the PIE community, and that any attempt on the part of the archaeologists
to reach a conclusion without due consideration of all the relevant linguistic data is
liable to lead them into serious error” (7). Linguistics must first do the groundwork:
“Once the spatial and temporal parameters of the putative homeland have been identi¬
fied on the basis of linguistic evidence, the archaeologist can set about the task of decid¬
ing which civilisation . . . can be plausibly associated with PIE” (7; italics in original).

In addition to this divide within disciplines that is evident from the disagreements
between archaeologists and archaeologists, and between linguists and linguists, and to
the cross-disciplinary divide between archaeologists and linguists, there is also a mas¬
sive East-West divide of which few in the West seem even aware. Until very recently,
Western scholars have paid little or no attention to, or are completely unaware of, the
reaction of scholars outside Western intellectual circles to the Indo-European debate.
Western scholars, whose primary emphasis and concern, at least historically, have been
the origins of Western civilization, have renegotiated and reconfigured the pre- and
protohistory of other nations such as India as by-products of their investigations. Yet,
for the most part, they have not been exposed to the concerns over, and responses to,
their formulations expressed by the native scholars from those very countries. India, in
particular, initiated the whole field of Indo-European studies when it’s language and
rich culture were “discovered” by Western scholars. Yet opinions from that country,
especially if in disagreement with the more forceful voices in the West, are poorly un¬
derstood or cursorily dismissed. This has deprived Western scholars of alternative views
that might force them to question their own inherited assumptions—assumptions that
have not always been shared by those outside mainstream Western academic spheres of
influence who have filtered the Indo-European problem through different historical,
religio-cultural, social, and political mind-sets. It has likewise deprived these Indian
scholars of valuable feedback from their Western peers. The following chapters will

Myths of Origin

attempt to help bridge this divide by articulating some of the more coherent, sober, and
rational responses to the Indo-European homeland problem that have been expressed
by intellectuals from the Indian subcontinent over the years. 19

As we turn to some of the reactions to this massive European intellectual enterprise,
it should not be surprising when some Indian scholars, trying to make sense of all this,
complain:

Instead of letting us know definitely and precisely where the so-called original home of
the Aryans lay, they drag us into a maze of conjectures clouded hy the haze of presump¬
tions. The whole subject of the Aryan problem is a farrago of linguistic speculations or
archaeological imaginations complicated hy racial prejudices and chauvinistic xenopho¬
bia. It is high time we extricate ourselves from this chaos of bias and belief. (Prakash,
1966, xliv)

Given the history of the Indo-European problem, it seems hardly surprising that many
Indian scholars have found themselves incapable of being co-opted by the prolific array
of theories produced by their Western colleagues. In a remarkably penetrating and well-
informed critique of the whole field written in the 1930s, when German Aryanism was
in full swing, B. N. Dutta (1936) states: “‘Germanism’ arose amidst the peculiar politi¬
cal condition of nineteenth century Germany. ... it has become the political shibboleth
of the occidental nations. ... we cannot see any reason why, in India, we should pin
our anthropological faith in [it]” (238). Criticizing the “slave psychology of the Indian
mind,” Dutta continues:

We find that pan-Germanic bias is in possession of the field of enquiry of the ancient
Indian civilization and Indian scholars are imbibing it through the medium of the En¬
glish language. In the field of anthropology, “Germanism” reigns supreme in India, the
Indians, . . . seeing the outside world only through the English language, have accepted
the views of the “Master” people as the only truth. . . . And we glory in it because it is the
gift of the “Master” people. (247)

Such comments do not necessarily reveal some antischolaTly quirk of a traditional
Hindu mind-set. On the contrary, I have referred to a history of cynicism in Western
intellectual circles over this issue. I hope, this chapter has suggested why, historically at
least, there might be some very good reasons for Indian scholars to be suspicious of the
whole enterprise, and of the ability of those engaged in it to make authoritative pro¬
nouncements on the early history of the Indian subcontinent and the Indo-Aryans. I
can perhaps set the stage and the tone for the next chapters by concluding this brief
survey of the history of the Indo-European homeland problem with some sympathy for
another disillusioned response voiced from somewhere in South India—one that is quite
representative of the Indigenous Aryan opinion of the whole Indo-European homeland
enterprise:

For nearly two centuries the investigations went on, and voluminous works were written
on the subject. The net result of their investigations ended in failure, and nothing defi¬
nite was settled either in the sphere of language or race. What they finally left behind is
the fiction of Ursprache [original language] with a false Urvolk [original people], who are
found located in an equally nebulous Urheimat [homeland]. (C. Pillai 1940, 2)

Early Indian Responses

It took considerable time for Indian literati to come to grips with the implications of the
Indo-European debates raging in Europe. Kopf (1969) has outlined the various avenues
through which European learning first became accessible to the Indian public. 1 There
was, of course, the nationalist response, which has been the element of most interest to
scholars studying Indian reactions to the colonial construction of ancient Indian his¬
tory. The nationalists were quick to incorporate Orientalist portrayals of ancient India
into their political agenda. As we have seen, the Orientalists, despite renegotiating cer¬
tain historical and temporal details, had shown genuine appreciation for the achieve¬
ments of the Aryan past, were quite happy to deter to the ancient Hindu Aryans as
more civilized and advanced than their ancient European Aryan contemporaries, and
generally contributed to a depiction of a previous golden age in which Hindus could
take vicarious pride.

As was the case in the West, diere were all sorts of reactions to, and appropriations
of, the discovery of a shared Aryan pedigree from the Indian subcontinent in popular,
political, and religious discourse. The first section of this chapter will briefly touch upon
nineteenth- and early twentieth-century nationalistic co-options of the Aryan theory in
terms of its applicability for Indian relations with the colonial power and for internal
power dynamics among competing sets of interests among Indians themselves. This
section could of course be the subject of a full treatment in its own right, so at the risk
of not doing sufficient justice to an important topic, I will extract a brief selection of
these reactions to provide something of a parallel to the Aryan discourse in Europe
during the nineteenth and early twentieth centuries. The socio-political context of the
modern period in India will be discussed in more depth in chapter 13. The second
section of this chapter will describe the first stirrings of opposition to the theory itself,
which were inaugurated by prominent religious leaders.

Early Indian Responses

Hindu Nationalist Responses

To many Hindus, the concept of Arya served primarily as a patriotic rallying cry.
Raychaudhuri (1988) outlines this immediate, and more euphoric, level of reflexive
popular response:

The Hindu self-image had received a moral boost from . . . the writings of Professor Max
Mueller. His linguistic studies stressed the common origin of Indo-European languages and
the Aryan races. These theories, translated into popular idiom, were taken to mean that the
master race and the subject population were descended from the same Aryan ancestors.

The result was a spate of Aryanism. Books, journals, societies rejoiced in the Aryan iden¬
tity. . . . Educated young men, in large numbers, affected a demonstrative reversion to the
ways of their forefathers—with fasts, pigtails, well-displayed sacred threads, and other stig¬
mata of Hindu orthodoxy. The name “Aryan” appeared in every possible and impossible
context—in the titles of books as much as in the names of drug stores. (34—35)

As outlined in the first chapter, Max Muller (1884) had been very influential in intro¬
ducing the theme of shared ancestry in India: “We recognize in Ram Mohan Roy’s
visit to England the meeting again of the two great branches of the Aryan race, after
they had been separated so long that they had lost all recollection of their common
origin, common language, and common faith” (11). Understandably, not all Hindus
were about to be taken in by this type of rhetoric; Muller himself quotes a “native
writer” from the Calcutta Indian Mirror (September 20, 1874) who exclaimed: “We
were niggers at one time. We now become brethren” (quoted in Chakrabarti, 1997,
99). Some had grown wary of Aryan discourses. But many Hindus, such as Tukaram
Tatya, took the opportunity to point out that “the difference between the European
and the Asiatic will be held to be of little moment” when consideration was directed
to the common Aryan bond. After all, since “the Hindus represent the older branch
of the great Aryan stock, . . . our European brethren should look upon us as filled
with the same blood” (93).

There have been a number of studies outlining the various nuances in the relation¬
ship between the Orientalist construction of the Aryan past and the Indian nationalist
movement (e.g., Leopold 1970). Scholars have long pointed out how early Orientalist
and Romantic themes such as “India was the cradle of the arts and sciences,” “Egypt,
Greece, and Rome were her pupils and recipients,” and “The Hindus were among the
first civilized nations when the nations of Europe had hardly risen above the hunting or
nomad state” were readily appropriated by Indian intellectuals, since they offered some
level of consolation to a subjected people (McCully 1966, 245-248). Moreover, Hindu
Aryanism could not just be vaunted as evidence of equality with the colonial rulers but
as proof of the Hindus’ moral superiority': British despotism and materialism were por¬
trayed as deviations from Aryan principles (Leopold 1 970, 278). Hindu reformers such
as Vivekananda (1970-73) felt that it was the western Aryans that were being given the
opportunity to learn from their Hindu Aryan brethren (i.e., more specifically, from him¬
self): “Which of us ever dreamt that a descendant of the old Indian Aryans, by dint of
Tapas, would prove to the learned people of England and America the superiority of
the ancient Indian religion over other creeds?” (3:350). Nor was this exchange just to
take place on Indian soil as a result of Western initiative; if the Western Aryans had

48 The Quest for the Origin of Vedic Culture

overpowered India materially, the Indian Aryans were destined to conquer the West
spiritually:

Two curious nations there have been—sprung of the same race . . . the ancient Hindu
and the Greek. The Indian Aryan . . . became introspective. The analysis of his own
mind was the great theme of the Indo-Aryan. With the Greek, on the other hand, ... his
mind naturally went outside. It wanted to analyze the external world. . . . Today the an¬
cient Greek is meeting the ancient Hindu on the soil of India. . . . We must be always
ready to sit at the feet of all. ... At the same time we must not forget that we have also
to teach a great lesson to the world. . . . the gift of India is the gift of religion and philoso¬
phy, and wisdom and spirituality. ... we must go out, must conquer the world through
our spirituality and philosophy. (Vivekananda 1970-73, 3:269-273) 2

Such sentiments were typical of the time. The Theosophists like Olcott also contrib¬
uted to notions of Hindu Aryan superiority in their addresses to groups such as the
Arya Samaj: “Recognizing as we do the Aryan source of our race and of its knowledge
of things terrestrial and celestial, we, Theosophists will feel proud to he permitted to call
ourselves your disciples” (Sarda 1946, 529; italics in original). Keshub Chandra Sen ([1901 —
4] 1954) later echoed similar themes when recognizing that “in the advent of the En¬
glish nation in India, we see a reunion of parted cousins, the descendants of two differ¬
ent families of the Aryan race.” Each had a valid role to play: “India in her present
fallen condition seems destined to sit at the feet of England for many long years, to
learn Western art and science. And, on the other hand, behold England sits at the feet
of hoary-headed India to study the ancient literature of this country” (325). Unlike some
of his other religious contemporaries, Sen did not hesitate to stress the duties Aryan
kinship involved that were incumbent on the materialistic side of the family: “May
England . . . [give] us as much of the light of the West as lies in her power! That is her
mission in India. May she fulfill it nobly and honourably. Let England give us her in¬
dustry and arts, her exact sciences and her practical philosophy” (325-326). His brother
was equally idealistic: “The Hindu and the Englishman are brothers! . . . every brother
man is learning to recognize in the face of his fellow-creatures the image of his first
forefathers. . . . Let that unity be the groundwork of future peace and brotherhood”
(Leopold, 273).

Not all were prepared to acknowledge the material advantages that might be gained
from the English Aryan brethren, however. Although K. C. Sen (1954) had waxed elo¬
quent about the benefits derivable from gallant Aryan England—“Fallen [Aryan] India cried
for help, and lo! at Heaven’s bidding England hastened to her rescue” (126)—others saw
things differently. The very first line of Lajpat Rai’s book England’s Debt to India (191 7) is
“India once was rich.” In contrast to Sen’s rhapsody, despite appropriating Orientalist
tropes of previous golden ages “when Greece and Italy, those cradles of European civiliza¬
tion, nursed only the tenants of a wilderness [and] India was the seat of wealth and gran¬
deur” (4), Rai was adamant that the British conquest of India had been “the most insidi¬
ous, most prolonged and most devastating to the conquered” (319). In this narrative,
England had plundered her Aryan sibling, not “hastened to her rescue.”

The more moderate Gokhale (1920), who was prepared to allow that other members
of the Aryan family “brought their own treasure into the common stock” (1023), also
appropriated Orientalist discourse: “The people of India are an ancient race who had
attained a high degree of civilization long before the ancestors of European nations

Early Indian Responses

understood what civilization was. India has long been the birthplace of great religions.
She was also the cradle and long home of literature and philosophy, of science and
arts” (925). Dayananda Saraswati refused to recognize any Hindu Aryan debt to Europe
even on a material level—everything came from India: “The people of Egypt, Greece or
the continent of Europe were without a trace of learning before the spread of Knowl¬
edge from India” (238).

The construction of a golden past is hardly unique to India. The development of
nationalisms almost invariably involves the creation of a sense of continuity between
the past and the present. This past is mined for material with which to construe a sense
of historic identity, unity, glory, and continuity to inspire political action in the present—
Hobsbawn’s “invention of tradition.” Obviously, these themes offer hope for a future
return to an idyllic state once real-life political obstacles are surmounted by adoption of
a nationalist agenda. Bipan Chandra (1984) has argued that, on the one side, the na¬
tionalist leaders needed a theme with instant psychological appeal that could inculcate
the idea of nationalism in the masses without alarming the imperialist powers; on the
other, the British encouraged this sense of identification with an idyllic spiritual Hindu
past so that the de facto material British present would not be jeopardized.

Just as the Aryan connection was configured to support a wide variety of domestic
and colonial agendas by Europeans, it surfaced in a variety of ways among Indians in
their internal negotiations with each other, in addition to their dealings with the exter¬
nal imperial power. The Aryan-Dravidian dichotomy was put to political use both by
Brahman elitists in the North and by Tamilian separatist voices in the South who were
quick to capitalize on the idea of Aryan invasions. From the former camp, for example,
Ranade approved of the derogatory descriptions made by Western scholars like Abbe
Dubois of the abominable practices extant in the South. In his view this situation oc¬
curred because Aryan Brahman influence had “hardly penetrated below the upper classes.”
The Aryan Brahmanical settlers were “too few in numbers and too weak in power to
make any lasting impression beyond their own limited circle upon the multitudes who
constituted the aboriginal races in the Southern Peninsula” (Ranade [1915] 1992, 205).
The Orientalist view that Hinduism consisted of the morally and culturally superior
Aryans who were detrimentally influenced by their merger with the backward and primi¬
tive aboriginals was happily regurgitated by many Brahmanas for whom Brahmanical
Aryanism corresponded to civilization.

Although not all Hindu nationalists participated in the denigration of Dravidian
culture, most did share a strong conviction that India could be saved by returning to
the purity of a reconstituted Sanskritic Aryanism. Ramaswamy (1997) and Irschick (1971)
outline the reaction to such attitudes that took firm root in the South in the form of
neo-Saivism. According to spokesmen from southern castes like the Chetti and Vellala,
who were particularly dismayed by the prospect of Brahmanical culture highjacking the
emerging nation, “it was not the Dravidians who corrupted a pristine Hinduism. . . .
on the contrary, it was Brahmanism and Aryanism that had debased the original Tamil
religion and diverted it from its hallowed path of monotheism, rationalism, and egali¬
tarianism into the ‘gutters’ of polytheism, irrational rituals, and unjust social hierar¬
chies” (Ramaswamy 1997, 29-30). For them the Dravidian religion far predated that of
the Aryans, not just in the South, but all over the subcontinent. Siva was a pre-Aryan
Tamilian deity whom the later Aryan intruders had pressed into service in their own

50 The Quest for the Origin of Vedic Culture

pantheon. By the time neo-Saivism was in full swing in the 1920s, it was not Sanskrit
but Tamil that was the world’s original, divine language. Others went further: “Most of
what is ignorantly called Aryan Philosophy, Aryan Civilization is literally Dravidian or
Tamilian at bottom” (Sundaram Pillai, quoted in Irschick, 1971, 152).

Phule, at the end of the nineteenth century, was one of die earliest proponents of
such ideas:

The aboriginals like the Gon ds and the Bhils were the masters (rulers) of this land India,
and the Iranians (Aryans) came to India at a later date (as invaders and interlopers). . . .
The Aryan invaders (Brahmins) desecrated the sacred sacrifices here, robbed and oppressed
the original inhabitants and stigmatised them as “Dasyus. . . . The (Indian) Civil Service
has been (unjustly) monopolized by the Aryan Brahmans (here) and I beg to submit that
it portends a great danger to the whole nation.” (Patil 1991, 132)

Others went further. Perhaps the best known detractor of Aryan culture was Periyar
E. V. Ramaswami, the leader of the Dravidian movement in South India. Ramaswami
despised almost everything that has come to be known as “Hinduism,” portraying it
as an Aryan imposition on an indigenous Dravidian populace. He exhorted that “the
Tamil . . . may liberate himself from the Aryan yoke.” In his version of things, “the
Aryans, when they invaded the ancient land of the Dravidas, maltreated and dishonoured
the latter and had written a false and coloured history wholly fallacious. It is this they
call Ramayana wherein Rama and his accomplishes are styled as Aryas, Ravana as
Rakshasa.” Ramaswami’s book is dedicated “to mirror to the Tamils what ascendancy
is given to the Aryan and how disgracefully the other communities are deprecated”
(Ramaswami, 1981, 2-3). It is Ravana, in Ramaswami’s reading of the plot, who is the
true Dravidian hero who attempted, unsuccessfully, to save his people from the exploi¬
tation and tyranny of the invading Aryans. Ramaswami’s mission was dedicated to
detaching, both culturally and politically, the life of his fellow Tamils from Brahman-
dominated Aryan influence. 3

Ambedkar also attempted to uplift those who had suffered the most at the hands
of Brahmanical Aryan culture: the Sudras. But, unlike Ramaswami, his method was
not to attempt to uncouple this social class from an alien Aryan culture. On the con¬
trary, according to Ambedkar (1946), “the Shudras were one of the Aryan communi¬
ties of the Solar race. . . . The Shudras did not form a separate Varna. They ranked as
part of the Kshatriya Varna in the Indo-Aryan society” (v). On the basis of a variety
of passages, particularly Mahabharata, santi parvan 38-40 (which describes a Sudra
by the name of Paijavana performing a major sacrifice conducted by Brahamanas),
Ambedkar argued that the Sudras were once wealthy, glorified and respected by rsis,
composers of Vedic hymns, and performers of sacrifice. Due to continuous feuding
with the Brahmana class, the Sudras inflicted many tyrannies on the Brahmanas, who,
in retaliation, denied them the upanayana initiation ceremony, causing them to even¬
tually become socially degraded.

Ambedkar, in his book Who Were the Sudras? (1946) offers a critique of the philo¬
logical basis of the Aryan invasion theory, that in places is well-informed and well-
argued. He adamantly rejected this theory, which he saw as partly responsible for propa¬
gating the erroneous idea that the Sudras were a non-Aryan, indigenous ethnic group.
Nonetheless, he does resonate with Periyar Ramaswami on one issue:

Early Indian Responses 51

The Aryan race theory is so absurd that it ought to have been dead long ago. But far from
being dead, the theory has a considerable hold upon the people. . . . The first explana¬
tion is to be found in the support which the theory receives from Brahmin scholars. This
is a very strange phenomenon. As Hindus, they should ordinarily show a dislike for the
Aryan theory with its express avowal of the superiority of the European races over the
Asiatic races. But the Brahmin scholar has not only no such aversion but most willingly
hails it. The reasons are obvious. The Brahmin believes in the two nation theory. He
claims to be the representative of the Aryan race, and he regards the rest of the Hindus
as descendants of the non-Aryans. The theory helps him establish his kinship with the
European races and share their arrogance and their superiority. ... it helps him main¬
tain and justify his overlordship over the non-Brahmins. (76)

Other spokesmen for the most disadvantaged castes had different ideas about how
to redress injustices. In their estimation, better gains might be had by accepting the
Aryan invasion theory, with all its implications, rather than rejecting it: “Even the present
Swarajists [those demanding independence]—the Aryans—were themselves invaders like
the Muhammadans and the Europeans. If this country has to be governed by aborigi¬
nes, all the offices must necessarily be filled by the original inhabitants—the Chamars,
the Kurumbas, the Bhils, the Panchamas, etc.” (quoted in Irschick, 1971, 154). As far
as some in the South were concerned, it was a “misrepresentation to say that the Brah¬
mins belong to the same Indian nation as the non-Brahmins while the English are
aliens. . . . Indian Brahmins are more alien to us than Englishmen” (Raghavan, quoted
in Irschick, 1971, 158).

In short, although the excesses of Aryan ideology in Europe would be hard to sur¬
pass, the Indians themselves were not averse to attempting to extract political mileage
from the Aryan theme to support their own agendas. Indeed, in about 1920, one Visnu
Sakharam Pandit filed an immigration court case in America, claiming to be a Euro¬
pean. Since immigration was closed to Asiatics at that time, the ingenious fellow said
he could prove that he was a Brahman and therefore a fellow Aryan. The argument was
even entertained for a while, until a California court ruled that the Aryan invasion theory
was precisely that: just a theory, and therefore not citable as credible proof for immigra¬
tion purposes.

The First Reactions: Hindu Religious Leaders

Before moving on to an examination of the historical evidence, 1 would like to touch on
another dimension to the Aryan invasion problem here, in the context of early Indian
responses, before addressing it in a more general way in chapter 13. This involves is¬
sues of epistemology. After all, the Aryan invasion theory had significant implications
for traditional Hindu concepts of history. In traditional Sanskrit sources, the Aryans
are portrayed as the enlightened and cultured members of a spiritually advanced civili¬
zation, and the Vedas, sacred to millions of Hindus, have traditionally been accepted by
the orthodox as spiritual revelations, transmitted by generations of sages through the
ages since time immemorial. They contain no reference to a primitive, nomadic origin
outside of the subcontinent.

British scholars, particularly the utilitarians, enthusiastically expanding the scientific
frontiers of post-Enlightenment, colonial Europe, were in no mind to seriously consider

The Quest for the Origin of Vedic Culture

such propositions, especially after science and reason had forced them to relinquish the
earliest historical claims of their own biblical tradition: “The Brahmens are the most
audacious, and perhaps the most unskillful fabricators, with whom the annals of fable
have yet made us acquainted” (Mill [1820] 1975, 34). The remote forefathers of the
Indians, these scholars informed their South Asian subjects, were, as Emeneau (1954)
was to put it later, “nomadic, barbarous looters and cattle-raiders” (287), not gentle sages
teaching spiritual truths. Those very Vedas, they claimed, had been composed by fierce,
nomadic tribesmen from Europe or central Asia—the Indo-Europeans, Indo-Germans,
or Aryans—who had usurped the Indian subcontinent some time in the second millen¬
nium b.c.e. by force of arms, enslaving and exterminating those they encountered in
their way. Viewed from within the framework of the rational, nineteenth- and twentieth
century European mind-set, these Vedas were hymns of war and booty, and lusty invo¬
cations to anthropomorphic gods, not esoteric, divine revelations transmitted by enlight¬
ened beings from some imaginary golden age.

Fully committed to this rational and empirical worldview, and fortified by disci¬
plines such as linguistics, epigraphy, numismatics, archaeology, anthropology, philol¬
ogy, and a host of other ‘ologies’, European scholars presented a historical account of
ancient India that was radically different from the narratives that orthodox Hindus
had preserved for many centuries. Moreover, these new disciplines were entirely in¬
congruous with traditional epistemology, which was predominantly exegetical. All of
a sudden, foreigners such as the Greeks, despite their traditional role as mlecchds (for¬
eigners) and barbarians, became one of the only reliable sources for determining Indian
chronology. Just as unexpectedly, the dates of heterodox figures such as the Buddha
became the cornerstones of any attempt at historical reconstruction. A comparison of
the sacred, Vedic language with rude, mleccha tongues from outside the sacred Bharata
varsa resulted in the eternal Veda being demoted to a historical evolute from an even
earlier language, Proto-Indo-European—a language spoken by coarse, violent horsemen
from the barbarian lands far to the northwest of what was recognized by the sdstras as
the sacred Aryavarta. A terminus a quo was established for the eternal Veda, correspond¬
ing to the supposed arrival of the Indo-Aryan branch of these Indo-Europeans into India
around 1 200 b.c.e. These interpretations of India’s historical data by Europeans were
made public through such institutions as the Royal Asiatic Society of Bengal and quickly
became the standard version of ancient history taught in the schools and colleges that
soon began to proliferate in British India. Understandably, such reductionistic, philo-
logically derived depictions bruised religious sensitivities. According to Aurobindo
(1956):

In ancient times the Veda was revered as a sacred book of wisdom, a great mass of in¬
spired poetry, the work of Rishis, seers and sages. . . . Truth . . . not of an ordinary but
of a divine inspiration and source. Is this all legend and moonshine, or a groundless and
even nonsensical tradition? . . . The European scholars . . . went on to make their own
etymological explanation of the words, or build up their own conjectural meanings of the
Vedic verses and gave a new presentation often arbitrary and imaginative. What they sought
for in the Veda was the early history of India, its society, institutions, customs, a civilisation-
picture of the times. They invented the dreory based on the difference of languages of an
Aryan invasion from the north. . . . The Vedic religion was in this account only a wor¬
ship of Nature-Gods full of solar myths and consecrated by sacrifices and a sacrificial

Early Indian Responses 53

liturgy primitive enough in its ideas and contents, and it is these barbaric prayers that are
the much vaunted, haloed and apotheosized Veda, (i—iii)

Such Indological depictions of the Vedic times continue to aggravate religious Hindus
to this day. As described by Agrawal (1996):

For thousands of years Hindu society has looked upon the Vedas as the fountainhead of
all knowledge . . . and the mainstay of Hindu culture. . . . Never have our historical or
religious records questioned this tact. And now, suddenly, in the last century or so, it has
been propagated that the Vedas do not belong to the Hindus, they were the creation of a
barbaric horde of nomadic tribes who descended upon North India and destroyed an
advanced indigenous civilization. (3)

The first generations of Indians who undertook the challenge of mastering these
unfamiliar methods of scholarship did so under the patronage and auspices of the British
themselves; thus they were hardly in a position to challenge any of the conclusions being
produced. Most of those who did submit to such authority were completely co-opted by
the power of European intellectual prowess or dependent on the patronage of Euro¬
pean institutions, and so accepted the new version of things unquestioningly. For those
who were disposed to critique the colonial version, however, but who were outside the
pale of mainstream facilities, the new unfamiliar disciplines such as archaeology and
linguistics seemed formidable. This must have created a deep sense of frustration—espe¬
cially for religious Hindu intellectuals. After all, if die Vedas were being shown as not
even accurate with regards to mundane, verifiable, historical, and temporal matters such
as the chronology and homeland of the Aryans, why should they be trusted as reposi¬
tories of ontological knowledge? We can recall Jones’s and Maurice’s parallel concerns
regarding Old Testament historicity when they encountered Indian sources of knowl¬
edge (with, in this earlier period, Europe on the defensive): “Either the first chapters of
Genesis . . . are true, or the whole Fabrick of our national religion is wrong” (Jones
1788, 255), and “arguments principally founded on the high assumptions of the Brah¬
mins . . . could their extravagant claims be substantiated, have a direct tendency to over¬
turn the Mosaic system, and, with it, Christianity” (Maurice [1812] 1984, 22-23).

Similar concerns are evident whenever any traditional society encounters moder¬
nity. In nineteenth-century Bengal, one can sense the tension involved in maintain¬
ing both faith in scriptural validity and intellectual integrity in the Sri Krsna Samhitd.
This book was written in 1879 by Bhaktivinode Thakur, a High Court judge in
Jagganath Puri and father of Bhaktisiddhanta Sarasvati Thakur, the founder of the
Gaudiya Math of the Chaitanya sampradaya that was to become an influential sect in
eastern India and Vrindavan. Bhaktivinode’s reaction to the Aryan invasion theory,
as outlined by Shukavak Das (1996-97), is the earliest orthodox Hindu perspective
on Shastric historicism that I have uncovered. 4 Unable to refute the historical formu¬
lations of European scholarship, Bhaktivinode adopted the tools of modern critical
scholarship and, citing Western authorities such as Wilford, Pratt, Playfair, and Davis,
at least nominally accepted the proposition that the Aryans had indeed entered India
from the Northwest (although he negotiated with reason and argument his own date
of 4463 B.C.E. for their arrival). However, Bhaktivinode did not allow this historicization
to undermine the transcendence of the Vedic (and, more specifically, the Vaisnava
Bhagavata) dharma.

54 The Quest for the Origin of Vedic Culture

In the introduction to the Krsna Samhitd, Bhaktivinode states that scripture is of two
types: arthaprada, knowledge which deals with phenomenal matters such as history, phi¬
lology, linguistics, anthropology and archaeology, and so on; and paramarthaprada, knowl¬
edge that deals with transcendence. According to Bhaktivinode, arthaprada, even though
derived from the sacred scripture, is answerable to human scrutiny and analysis and
therefore can be adjusted and corrected according to time and place. Historical details
are negotiable. Paramarthaprada, in contrast, is inaccessible to human reason. It is tran¬
scendent and beyond the purview of human interpretation and speculation (although
realizable by the direct perception of the soul).

Bhaktivinode, in line with many other Hindu reformists of his day, concluded that
much in the sastras is aimed at attracting the neophyte religious consciousness by means
of superhuman stories and fantastic time calculations. In nineteenth-century British Ben¬
gal, however, such accounts, alienated many of die more intellectual and westernized
urban Bengalis from their religious traditions. The result was the creation of a spiritu¬
ally disenfranchised Hindu intelligentsia whose intellectual needs made them vulner¬
able to modern, Western ideas and who were likely to be dismissive of traditional reli¬
gious perspectives. This was Bhaktivinode’s target group. To attract the minds of this
educated but disoriented class to the paramarthaprada essence of the sastra, he was quite
willing to utilize critical analysis to negotiate the arthaprada portions, such as the ques¬
tion of Aryan origins, in accordance with the intellectually authoritative sources in his
day that were at his disposal. If the times were dominated by European methodologies
such as linguistics and archaeology, then Bhakivinode had no difficulty adopting these
methods and the conclusions they produced in order not to alienate those influenced
by this type of rationalism. As a spiritual leader, his concern was to retain his contem¬
poraries within the Vedic fold—specifically the Krsna-centered realm of the Chaitanya
tradition—while simultaneously encouraging them to engage intellectually with Western
critical thought.

As for his more traditional-minded colleagues, Bhaktivinode could only appeal to
them to try and understand the spirit of his historical formulations:

With folded hands I humbly submit to my respected readers, who hold traditional views,
that where my analysis opposes their long held beliefs, they should understand that . . .
what I have said about dharma applies to everyone, but with regard to matters which are
secondary to dharma, my conclusions are meant to produce benefits in the form of intel¬
lectual clarification only for qualified specialists. All the subjects which 1 have outlined in
the Introduction concerning time and history are based on the logical analysis of sastra,
and whether one accepts them or not does not affect the spiritual conclusions. History
and time are phenomenal subject matters and when they are analyzed according to sound
reasoning much good can be done for India. (Quoted in Shukavak Das, 1996-97, 139)

In this way, Bhaktivinode salvaged what he considered to be the essential aspects of the
Vedic and Bhagavata tradition from the firing line of any potentially embarrassing dis¬
coveries and conclusions of modern historical research. His absolute transcendence (the
saguna aspect of Sri Krsna) could then reside securely out of harm’s way, safe from is¬
sues of historicity. Bhaktivinode did, however, invite future scholars to reexamine and
improve the external, historical part of his formulations, such as the Aryan invasion,
when developments in the appropriate fields permitted.

Early Indian Responses 55

It took time for Indian scholars to adopt and learn the methods of Western critical
scholarship to the point where they felt comfortable enough to actually challenge the
status quo. The first outright voices of opposition against the Aryan invasion theory
were not raised until the end of the nineteenth century. They were inaugurated by promi¬
nent religious figures, whose discourses were in direct political and religious response
to the hegemony of British intellectual power, which was portrayed as untrustworthy in
the face of Vedic sabda pramana ‘scriptural evidence'. In 1882, Dayananda Saraswati,
founder of the Arya Samaj, protested:

In the face of these Vedic authorities how can sensible people believe in the imaginary
tales of the foreigners ... no Sanskrit book or history records that the Aryas came here
from Iran, and defeating the inhabitants of the country in battles, drove them away and
proclaimed themselves the rulers of the country. How can then the writings of foreigners
be worth believing in the teeth of this testimony? (266)

Elsewhere Dayananda drew on a literal reading of the Mahabharata to claim that “the
Aryas were the sovereign rulers of the whole earth” (329). But he did make some efforts
to familiarize himself with European scholarship and, in the 1860s, employed a Bengali
to read him Muller’s translation of the Rgveda. His choice of Tibet as the homeland of
the Aryans reflected the preference of Europeans such as Kant and Herder. Influenced
by the Orientalist critique in ways that partially paralleled (from an Aryan perspective)
the neo-Saivism of the South, his Samaj was dedicated to reestablishing the pristine
monotheistic purity of the early Aryan Vedic literature, and jettisoning the later polytheistic
accretions, superstitions, and corruptions drat had accrued during the Puranic period.

The Theosophists, who established their principal ashram in South India, retained
their belief in an Indian homeland well after such a position had long been considered
passe in Europe. Olcott, (1881), as noted earlier, considered that “the Brahmins have
their own chronology and no one has the means of proving that their calculations are
exaggerated. . . . We Europeans .. . have a right to more than suspect that India 8,000
years ago sent out a colony of emigrants” (124). Blavatsky ([1892] 1975) likewise stated
that “it has now become very clear to me that the Scandinavian, Egyptian, Greek, Cen¬
tral Asiatic, German and Slavonic gods were nearly all . . . born in prehistoric India”
(608). She also attempted to make a case for the antiquity of the Vedas. Juxtaposing
“the least age we can accord to the human race,” namely, 240,000 years, with Max Muller’s
statements that the Veda represents “the very infancy of humanity, and when hardly
out of its cradle,” she quips that “it really seems the duty of the eminent Sanskritist and
Lecturer on Comparative Theology to get out of this dilemma. Either the Rig-Veda hymns
were composed but 3,000 years ago, and, therefore, cannot be expressed in the ‘lan¬
guage of childhood’ ... or we have to ascribe to them an immense antiquity in order to
carry them back to the days of human mental antiquity” (Blavatsky n.d., 47).

Aurobindo (1971), who also kept abreast of European knowledge, expressed his
misgivings some years later: “The indications in the Veda on which this theory of a
recent Aryan invasion is built are very scanty in quantity and uncertain in signifi¬
cance. There is no actual mention of any such invasion. The distinction between Aryan
and un-Aryan on which so much has been built seems on the mass of evidence to
indicate a cultural rather than a racial difference” (24). This absence of any mention
in the Vedas of an external origin for the Aryans is the single most repeated objection

56 The Quest for the Origin of Vedic Culture

raised by Indian scholars—an issue to which I will return shortly. Vivekananda, (1970—
73), the first prominent Indian religious figure in the modern period to ignore the
prohibition of the sastra against crossing the seas, and the first to address an audi¬
ence in the West, not only rejected the Aryan invasion theory but also issued a rally¬
ing cry to scholars among his countrymen to oppose it:

And what your European pundits say about die Aryans swooping down from some for¬
eign land, snatching away the lands of the aborigines and setding in India by exterminat¬
ing them, is all pure nonsense, foolish talk! In what Veda, in what Sukta do you find that
the Aryans came into India from a foreign country? Where do you get the idea that they
slaughtered the wild aborigines? What do you gain by talking such nonsense? Strange
that our Indian scholars, too, say amen to them; and all these monstrous lies are being
taught to our boys! . . . Whenever the Europeans find an opportunity, they exterminate
the aborigines and settle down in ease and comfort on their lands; and therefore they
think the Aryans must have done the same! . . . But where is your proof? Guess work?
Then keep your fanciful ideas to yourself. I strongly protested against these ideas at the
Paris Congress. I have been talking with the Indian and European savants on the sub¬
ject, and hope to raise many objections to this theory in detail, when time permits. And
this I say to you—to our pundits—also, “You are learned men, hunt up your old books
and scriptures, please, and draw your own conclusions.” (5:534-535)

After Vivekananda’s death, some notes were found among his papers containing forty-
two points jotted down for a book he intended to write, ten of which dealt with issues
connected to the Aryan invasion theory. However, it remained for others to take up
Vivekananda’s stirring call to arms.

Conclusion

We have seen how Aryanism was coopted in all sorts of contradictory, but not necessary
conflicting, ways by missionaries, colonialists, Orientalists, anthropologists, nationalists,
and all manner of other ideologues in Europe throughout the nineteenth and twentieth
centuries. The Aryan connection was configured in support of just about any agenda. A
parallel situation holds true for India. Politically, in terms of its European connections,
Aryanism was welcomed by some as evidence of equality between the colonizers and the
colonized—and could even be invoked to elicit assistance from tire English Aryan brethren.
In other discourses, it was heralded as proof of the spiritual superiority of the Hindu Aryans
in comparison to the materialistic Western Aryans. In terms of its ramifications in die Indian
political scene, it was deferred to as proof of Brahmanical Aryan superiority over the rest of
the subcontinent or, contrarily, as proof of Brahmanical Aryan exploitation of the same.

From religious perspectives, however, the theory was more fundamentally troubling
at least from Brahmanical viewpoints. It completely undermined traditional concepts of
history and portrayed the wise Aryans of yore as little better than marauding barbar¬
ians. It was on the basis of the scriptural evidence that the first voices challenging the
basis of the theory—as opposed to coopting it for ideological purposes—were raised.
Suspicion of the theory based on scriptural testimony—or lack thereof—remains an ex¬
plicit or implicit factor in much Indigenous Aryan discourse. The science known as
philology in the West, was to provide the terrain for the first forays by Indian Sanskritists
against the conclusions of their Western peers. It is to this that I turn next.

Vedic Philology

Traditional Indian scholars immediately became suspicious of Western philology once
it began to conflict with traditional conceptions of the Veda and its origins. It was pri¬
marily through philology that the Indian homeland proposed by the earlier romantic
school was rejected by most scholars. 1 In 1909, the Imperial Gazeteer of India noted that
“the uniformity of the Indo-Aryan type can be accounted for only by one of two hypoth¬
eses—that its members were indigenous to the Punjab, or that they entered India. . . .
the opinion of European scholars ... is unanimous in favour of the foreign origin of
the Indo-Aryans. The arguments appealed to are mainly philological” (300).

Unlike comparative linguistics and archaeology, which were European innovations
and alien to Brahmanical thought, Vedic philology was an area much closer to tradi¬
tional Indian epistemological modes, since it required an expertise in Sanskrit and
involved the study of familiar traditional texts; Indian scholars were much better equipped,
in this discipline, to scrutinize the scholarship of Western savants. Although himself a
latecomer to Sanskrit studies, Aurobindo (1971) considered many of the conclusions of
the comparative philology of his day, despite containing “much that is useful,” to be
ultimately “an interesting diversion for an imaginative mind,” the fruits more of “an
ingenious play of the poetic imagination” (26). Aurobindo is reservedly appreciative but
not co-opted or intimidated:

Modern Philology is an immense advance on anything we have had before the nineteenth
century. It has introduced a spirit of order and method in place of mere phantasy; it has
given us more correct ideas of the morphology of language and of what is or is not pos¬
sible in etymology. It has established a few rules which govern the detrition of language
and guide us in the identification of the same word or of related words as they appear in
the changes of different but kindred tongues. Here, however, its achievements cease. The

58 The Quest for the Origin of Vedic Culture

high hopes which attended its birth have not been fulfilled by its maturity. There is, in
fact, no real certainty as yet in the obtained results of Philology. . . . Yesterday we were all
convinced that Varuna was identical with Ouranos, the Greek heaven; today this identity
is denounced to us as a philological error; tomorrow it may be rehabilitated. . . . We have
to recognize in fact that European scholarship in its dealings with the Veda has derived
an excessive prestige from its association in the popular mind with the march of Euro¬
pean science. The truth is that there is an enormous gulf between the patient, scrupulous
and exact physical sciences and these other brilliant, but immature branches of learning
upon which Vedic scholarship relies. (27-28)

Once the seemingly infallible facade of Oriental scholarship had been challenged by
the opposition of prominent religious figures, one of the first attempts to refute the
Aryan invasion theory by engaging the actual evidence at stake with reason and argu¬
ment was published in 1884 in the Theosophist. The contributor, Ramchandra Rao (1880),
voices his incredulity at the opinions of" his day:

We are told that the Aryan family, which lived in Central Asia, were a civilized people;
and that their religion was that of the Vedas. They had chariots, horses, ships, boats,
towns and fortified places before the separation took place. They were therefore not no¬
mads. Max Muller adds that the younger branch left first and emigrated into Europe. . . .
the oldest quitted its ancestral abode last of all, for a new home in India. The inference
to be drawn, then, is that the old home was abandoned by every soul, and left to become
a dreary and a desolate place as we now find it. . . . the efforts of philology . . . can hardly
succeed in metamorphosing a vague theory into real Simon Pure, but must remain as
they are—a hollow farce. (306-307)

In his opinion, the whole Aryan invasion theory was “nothing but a varnished tale,
utterly undeserving of the name of traditional history” (305).

Similar misgivings were voiced in 1901 by one Aghorechandra Chattopadhyaya in
Calcutta. In this book, one can sense the author seriously struggling to make sense of
the conclusions of Western scholarship, yet unable to conceal his own bewilderment at
the theories that he was encountering; “Whatever might be the credibility the scholars
are blessed with, we can hardly reconcile ourselves with such an easy faith on a matter
like this.” Commenting on the spectacular achievements of the subbranches of the Indo-
European family, such as the Vedic Indians, Greeks, Romans, and Persians, he won¬
dered, with remarkable acumen for his time and sources, how the main trunk of the
Indo-European tree could have produced such conspicuous fruits that survived for mil¬
lennia, and yet leave no trace of itself:

While the major branches of the main trunk gathered strength, looked healthy, and
spread far and wide, the latter, at the same time, withered, shriveled, and failed to show
any indication of life and vitality and disappeared from sight and was lost for ever with¬
out leaving any trace or mark that might lead to its identification, nor could any fossil
remains of it be detected or found out, so that it could be inferred that such a society
in such a stage of development existed at one time, on the surface of the earth. ... A
story so imperfect in every important respect is put forward seriously for people to be¬
lieve in and accept as an authentic account of the ancient history of the Indo-European
race. (59)

Chattopadhyaya also struggles to make sense of what appeared to him to be the con¬
tradictory proposals that the Indo-Europeans were wandering nomads and yet were

Vedic Philology 59

held to have originated from a specific abode, and that they were primitive tribesmen
and yet were able to formulate and utilize a language as intricate and complex as Indo-
European. Albeit on a rudimentary level, some Indian scholars were beginning to
pay closer attention to the specifics underlying the philological theories and specula¬
tions of Western scholars. There were two specific philological areas drat were fundamental
in asserting an Aryan invasion of India: the racial evidence, involving the distinc¬
tions between the Aryas and their enemies the Dasas, and the geographic parameters
of the texts themselves.

The Racial Evidence

The first prominent note of discord between traditional exegesis and Western scholar¬
ship was sounded because of the lack of explicit mention, in the Vedic texts, of a for¬
eign homeland of the Aryan people. As mentioned previously, this conspicuous silence
had been noted even by nineteenth-cenmry Western scholars (e.g., Elphinstone 1841).
The absence of any mention of external Aryan origins in traditional Sanskrit sources is,
to this day, perhaps the single most prominent objection raised by much of the schol¬
arship claiming indigenous origins for the Aryan culture. This consideration was summed
up succinctly by Srinivas Iyengar in 1914:

The Aryas do not refer to any foreign country as their original home, do not refer to
themselves as coming from beyond India, do not name any place in India after the names
of places in their original land as conquerors and colonizers always do, but speak of them¬
selves exactly as sons of the soil would do. If they had been foreign invaders, it would
have been humanly impossible for all memory of such invasion to have been utterly
obliterated from memory in such a short time as represents the differences between the
Vedic and Avestan dialects. (79-80)

A few Western scholars had tried to find some oblique references or reminiscences of
the pre-Vedic people during their trajectory over central Asia. In 1913, Hillebrandt found
reason to suppose that the Hariyupiya in RV 6.27.8 “is the Ariob or Haliab, a source
river of the Kurum” (49). Other attempts to find traces of the Indo-Aryans in Iran and
other places outside of India will be discussed in chapter 7. One should also note that
several other Indo-European cultures, such as those of the Greeks and Scandinavians,
also preserve no mention of their migrations into their historical territories, yet we know
they were immigrants there at some point. That a historical event is lost from the collec¬
tive consciousness of a people due to the passage of time, does not indicate that the
event never took place. For the present purposes, the fact that the Vedas themselves
make no mention of any Aryan invasion or immigration reveals a major epistemologi¬
cal concern in this debate. Scriptural testimony, sabda pramana, in varying degrees, still
holds a preeminent status as an authoritative source of historical information in the
view of many Indian scholars.

Once the warning alarm had been raised regarding the lack of explicit mention of
Aryan invasions, scholars began to look more carefully at the implicit evidence Western
scholars had brought forward in this regard. It was the racial interpretations imposed
on various Vedic passages, particularly those that referred to the battles between the
Aryans and their foes, the Dasas or Dasyus, that aroused the indignation of Indian

60 The Quest for the Origin of Vedic Culture

scholars. Aurobindo (1971), again, was an outspoken, witty, and penetrating forerunner
in this regard:

It is urged that the Dasyus are described as black of skin and noseless in opposition to
the fair and high-nosed Aryans. But the former distinction is certainly applied to the Aryan
Gods and the Dasa Powers in the sense of light and darkness, and the word anasah does
not mean noseless. Even if it did, it would be wholly inapplicable to the Dravidian races;
for the Southern nose can give as good an account of itself as any “Aryan” proboscis in
the North. (24)

The racial interpretations of the Vedic passages were inaugurated by Max Muller, who
is both the hero and the archfiend of the Indigenous Aryan school. Factually: “The first
effort to find direct evidence of the physical features of the Indian aborigines in the
Sanskrit texts dating from the time of the Big Bang that brought Indian civilization into
existence . . . boiled down to a matter of noses” (Trautmann, 1997, 197).

Muller (1854b), searching for clues in the Rgveda that might provide evidence of
this “big bang,” decided: “The only expression that might be interpreted in this way is
that of ‘susipra,’ as applied to Aryan gods. It means ‘with a beautiful nose’ . . . The
Dasa or barbarian is also called vrsasipra in the Veda, which seems to mean goat or
bull-nosed, and the ‘Anasas’ enemies whom Indra killed with his weapon (RV V,29,10)
are probably meant for noseless . . . people”. (346). Muller later recanted his interpre¬
tation of the word s'ipra, so the evidence was reduced to a solitary word, andsa, in a
single passage. This sole possible description of the Dasa nose, however, like Pinocchio’s
nasal organ, was to have an expanded life of its own. By 1891, H. H. Risley, who was
compiling his ethnological material on Indian tribes and castes, was able to say that
“no one can have glanced at the Vedic accounts of the Aryan advance without being
struck by the frequent references to the noses of the people whom the Aryans found in
possession of the plains of India [whom] they spoke of as ‘the noseless ones’” (249-
250; my italics). The solitary nasal reference had suddenly become a frequent one.

McDonnell and Keith (1967), while at least acknowledging that both the pada text and
Sayana, the oldest existing commentator on the Rgveda, had interpreted the word andsa
as meaning the equally valid alternative translation ‘without face’ (which is how Geldner
and Grassman had accepted it) as opposed to ‘without nose’, nonetheless further cemented
Muller’s identification with their approval. As far as they were concerned, it “would ac¬
cord well with the flat-nosed aborigines of the Dravidian type, whose language still per¬
sists among the Brahuis, who are found in the North-West” (348). Muller had construed
the word as a-nasa, ‘without nose’, as opposed to an-as ‘without mouth or face’, as Sayana
had construed it. The word occurs in a passage where the Dasyus are also described as
mr dhavacah, which is glossed by Sayana with himsitavagindriyan ‘having defective organs
of speech’. This could reasonably simply refer to people considered rude or uncultivated
barbarians by their Aryan detractors rather than to any racial term. However, the quest for
textual evidence of the Aryan invasion caused the racial interpretation to be favored, and
it is this interpretation that has continued to surface up to the present day: “The Vedas
recognize a dichotomy between the Indo-Aryans and their dark-skinned enemies, tire Dasa,
who are on one occasion described as ‘nose-less,’ which has generally been interpreted as
a pejorative reference to Dravidian physical features” (Mallory 1989, 45).

Vedic Philology 61

Srinivas Iyengar, in 1914, was not convinced by this type of “great scientific hardi¬
hood”:

One solitary word anasa applied to the Dasyu has been quoted by . . . Max Muller . . .
among numerous writers, to prove that the Dasyus were a flat nosed people, and that,
therefore, by contrast, the Aryas were straight-nosed. Indian commentators have explained
this word to mean an-asa, mouthless, devoid of fair speech. ... to hang such a weight of
inference as the invasion and conquest of India by the straight nosed Aryans on the solitary
word andsa does certainly seem not a very reasonable procedure. (6)

Iyengar is equally unimpressed by the racial interpretations of other passages in the

Veda that had been given by Western scholars:

The only other trace of racial reference in the Vedic hymns is the occurrence of two words,
one krishna in seven passages and the other asikini in two passages. One of the meanings
of these two words is “black,” but in all the passages, the words have been interpreted as
referring to black demons, black clouds, a demon whose name was Krishna, or the pow¬
ers of darkness. Hence to take this as evidence to prove that the invading Aryans were
fair-complexioned as they referred to their demon foes or perhaps human enemies as
black is again to stretch many points in behalf of a preconceived dreory. (6-7)

Iyengar is well worth quoting at length, because his arguments are well researched and

penetrating:

The word . . . Arya occurs about 33 times iin the Rgveda]. . . . the word Ddsa occurs
about 50 times and Dasyu about 70 times. . . . The word Arya occurs 22 times in hymns
to Indra and six times in hymns to Agni, and Dasa 50 times in hymns to Indra and twice
in hymns to Agni, and Dasyu 50 times in hymns to Indra and 9 times in hymns to Agni.
The constant association of these words with Indra clearly proves that Arya meant a
worshipper of Indra (and Agni). . . . The Aryas offered oblations to Indra. . . . The Dasyus
or Dasas were those who were opposed to the Indra Agni cult and are explicitly described
thus in those passages where human Dasyus are clearly meant. They are avrata without
(the Arya) rites, anyavrata of different rites, ayajavdna, non-sacrificers, abrahma without
prayers, also not having Brahmana priests, anrichah without Riks, brahmadvisha, haters
of prayers to Brahmanas, and anindra without Indra, despisers of Indra. They pour no
milky draughts, they heat no cauldron. They give no gifts to the Brahmana. . . . Their
worship was but enchantment, sorcery, unlike the sacred law of fire-worship, wiles and
magic. In all this we hear but the echo of a war of rite with rite, cult with cult and not one
of race with race. (5-6) 2

Others have voiced just as penetrating critiques:

In the attempt to ransack the latter-day Sanskrit texts for proofs of Nordic characteris¬
tics, ... we forget that if in latter day Sanskrit texts sentences such as “Gaura [white,
yellowish], . . . pingala [reddish brown, tawny, golden], kapilkesa [brown or tawny hair]”
are to be found in Patanjali’s Mahabhasya (V. 1. 115) and if Manu has said that a Brahmana
should not marry a girl with pingala hair (38) there are other sentences in previous ages
which contradict the strength of these characteristics. But with the help of these two
sentences attempt is being made to prove the existence of Nordic characteristics amongst
the Indian people. . . . The God Rudra is described to have possessed golden hair . . .
yet we cannot make a Nordic viking out of him, as he had brown-hued skin-colour and
golden-coloured arm. . . . Surely we cannot take the god Rudra as a specimen of race-

The Quest for the Origin of Vedic Culture

miscegenation. ... we beg to state that these allegories should be accepted as poetic
fancies. They cannot be used as scientific data, for anthropological purpose. (Dutta 1936,
248-252)

Interestingly, almost a full century after Indian scholars started objecting to the racial
interpretations imposed on die Arya-Dasa dichotomy, Western scholars have recently
also started drawing attention to nineteenth-century philological excesses. Levitt (1989),
in his analysis of the word andsa, points out that even if it does mean ‘noseless’, an
equivalent term in the language of the Bhil tribe is used in an ethical as opposed to a
racial sense to indicate someone who is untrustworthy. Schetelich (1990), in turn, has
analyzed the three occurrences of the phrase Icrsna (or asikni) tvac used in conjunction
with the dasyu, which has generally been translated as ‘dark skin’ (247). Her conclu¬
sion is that the word is a symbolic expression for darkness. Witzel comments on the
same term that “while it would be easy to assume reference to skin colour, this would
go against the spirit of the hymns: for Vedic poets, black always signifies evil, and any
other meaning would be secondary in these contexts” (1995b, 325, fn). These realiza¬
tions can be found in any number of Indigenous Aryan publications stretching back for
at least a century. Trautmann (1997), in his analysis of the development of, and inter¬
action between, ethnology and philology in the nineteenth century, finds his own experi¬
ment of subjecting the evidence for the racial interpretation of Indian civilization to a
minimizing reading revealed just how soft that evidence is and the amount of overreading
upon which it is based (213). He points out that the racial theory of Indian civilization
is the product of the late nineteenth century, when the relations between whites and
other ethnic groups in the Anglo-Saxon world were being reconfigured with ideological
support from a spate of racial essentialism (208). Trautmann concludes: “That the racial
theory of Indian civilization still lingers is a miracle of faith. Is it not time we did away
with it?” (215).

Hock (1999b) suggests that the reluctance to review this racial material is due to the
failure to live up to the scholarly ideals of constantly reexamining the evidence. He, too,
undertook a similar exercise by extracting all the passages that Geldner had construed
in a racial sense in his translation of the Rgveda and found them all to be either mis¬
translated or, at least, open to alternative nonracial interpretation. The reason racial
readings were preferred was due to the “quasi-scientific attempts to provide a justifica¬
tion for ‘racially’ based European imperialism. . . . Moreover, the British take-over of
India seemed to provide a perfect parallel to the assumed take-over of prehistoric India
by the invading Indo-Aryans” (1999b, 168).

It seems fair to note, however, that for over a century many Indian scholars have
been aware of, and objected to, such biased readings all along: “Thus ‘Arya’ moved
from the Vedic literature to the European political arena. . . . They thought that as the
Vedic people were the most cultured people of antiquity, they cannot but be ‘white men,’
no matter whether blonde or brunette, who conquered the noseless dark people of the
Indus Valley” (Chandra 1980, 123). 3 B. R. Ambedkar (1946) delivered a particularly
scathing critique of the whole enterprise of attempting to establish invasions on the
basis of racial evidence in the Rgveda and concludes:

Why has the theory failed? . . . The theory of an invasion is an invention. This invention
is necessary because of a gratuitous assumption which underlies the Western theory. The

Vedic Philology 63

. . . assumption is that the Aryans were a superior race. This theory has its origin in the
belief that the Aryans are a European race and as a European race it is presumed to be
superior to the Asiatic ones. . . . Knowing that nothing can prove the superiority of the
Aryan race better than invasion and conquest of the native races, the Western writers
have proceeded to invent the story of die invasion of India by the Aryans, and the con¬
quest by them of the Dasas and Dasyus. . . . The originators of the Aryan race theory are
so eager to establish their case that they have no patience to see what absurdities they
land themselves in. They start on a mission to prove what they want to prove and do not
hesitate to pick such evidence from the Vedas as diey think is good for them. (72—7.5)

Of course, despite offering elaborate and, in places, well-argued and legitimate refuta¬
tions of the racial evidence along some of the lines outlined here, Ambedkar’s research
was not without a clear, and philologically questionable, agenda of its own, as was noted
in chapter 2. But his point here holds good.

Philologists are not alone in being unable to identify any compelling racial traits in
the Rgvedad Present-day archaeologists also concur that there are no innovations in the
skeletal remains of humans found in the subcontinent that necessarily correspond to
an incoming group of people that are in any way distinct from a separate indigenous
group of people. This evidence will be discussed in chapter 11. In terms of the literary
material, in addition to the so-called racial references, another body of philological evi¬
dence has been very influential in supporting the position that the Aryans were immi¬
grants into the subcontinent. This is based on the geographic boundaries alluded to in
the texts themselves.

The West-toEast Geographic Shift in Sanskrit Texts

In 1860, Muir, in his arguments raised against the consideration of the Aryans being
indigenous to India that was still lingering in his time, was the first scholar to attempt
to argue extensively that the Sanskrit texts themselves could be used to demonstrate an
Aryan invasion of India. Although beginning his thesis with a “candid admission” that
“none of the Sanskrit books, not even the most ancient, contain any distinct reference
or allusion to the foreign origin of the Indians” (Muir [1860] 1874, 322), he developed
his case by documenting how the geographic horizons referred to in progressively later
Sanskrit texts expand from the northwest part of the subcontinent to the eastern and,
eventually, the southern parts. From this perspective, this textual awareness of increas¬
ing portions of India corresponded to the actual physical expansion of the Indo-Aryans
themselves into India from the Northwest and dien across the subcontinent.

Muir’s arguments began with the Rgveda, which refers to the Kubha river in
Afghanistan and is firmly situated in the Punjab between the Indus and the Sarasvatl:
“The oldest hymns of the Veda show us the Arian people still dwelling . . . between dre
Cabul river and the Indus” (339). 3 However, “the Ganga and the Yamuna are only
mentioned once in the tenth book,” and the southern Vindhya mountain range is not
mentioned at all (347). This suggested to him drat the Aryans had just begun to move
east by the time they were compiling the later books of the Rgveda (such as the tenth).
Muir fortified his case by interpreting the conflict between the Aryas and the Dasyus in
the racial manner discussed above, and then quoted a passage from the Satapatha

64 The Quest for the Origin of Vedic Culture

Brahmana wherein “the gradual advance of the Aryas with their Brahmanical worship,
from the banks of the Sarasvati eastward to those of the Sadanira, and afterwards be¬
yond that stream, is . . . distinctly indicated” (403). 6 Finally, the Ramayana, with the
conquest of Lanka, completed the Aryan expansion across tire Vindhyas and on to the
South. The fabulous creatures and beings described in this Epic are representations of
the Aryan encounter with the aborigines—a proposal Muir bolstered with Ellis and
Campbell’s discovery that the Dravidian languages belonged to a different family from
those of the North. Other geographic references have been brought forward since Muir
to solidify this basic line of argument.

Muir’s logic, then, is that certain parts of India are not mentioned in the oldest part
of the Rgveda because the Aryans had not yet been to those parts. This argument can
be used both ways, however. If it is to be deduced that because the Rgveda does not
explicitly mention the East and the South of India, then the Aryans had not yet been or
gone to these regions, then the same parallel logic cannot be denied for the absence of
any explicit mention of the Caspian Sea or its environs in the Rgveda: one would be
hard-pressed to use these particular grounds to disallow the oft-cited claim that the Aryans
had equally not been, or come from, those regions either, since they too are not explic¬
itly mentioned in any texts. If Vedic geography is silent regarding eastern and southern
Indian landscapes and peoples, it is arguably also ignorant of distant, external north¬
western landscapes and peoples. From this perspective, the two lacunae, both merely
argumenti ex silentio, seem to negate each other in terms of providing solid evidence of
the migrations of these people.

One line of argument attempts to brush off these geographic parameters by stating
that all that the evidence indicates, without making assumptions, is that die composers
of the hymns, who were certainly not cartographers, happened to make some peripheral
references to the immediate areas where they lived and happened to make no mention
of other places that were irrelevant to their hymns. Some scholars do not feel obliged to
conclude front this that the Indo-Aryans were ignorant of other places because they had
not yet arrived farther east or south in the subcontinent. The composers of later Vedic
texts, and of the Epics and Puranas, happened to live elsewhere and therefore described
different locations connected with the themes and events they were concerned with:

What does it prove at best! It only proves that the people who sang the hymns lived in
the land of the “Five (or Seven) Rivers”; nothing beyond that. It does not, for instance,
prove that the land beyond the “Five Rivers” was not inhabited. ... [In the] Ydjurvedic
tradition . . . Yajnavalkya [was] an inhabitant of Mithild, which is very far removed from
the “Five Rivers”. ... If then, at least one of the Vedic Seers inhabited the Eastern land
of Mithila, and some inhabited the land of the Five Rivers—what definite conclusion can
that lead to? (Jha 1940, 2)

This type of dismissal, however, needs to address the detailed philological work of Witzel
(1989) who not only has excavated tire geographic horizons known to each separate
stratum of the Vedic texts but also has attempted to identify the different waves of tribal
units that inaugurated the expansions and even the dialectal variants and archaeologi¬
cal cultures that accompanied the spread. He, too, notes that the Rgveda is limited to
the Punjab and it surroundings, while the Atharvaveda knows all of the North Indian
plains of the Ganges-Yamuna doab of Uttar Pradesh. The late Brahmana texts, in con-

Vedic Philology 65

trast, “suddenly have a geographical horizon reaching from Gandhara (and beyond) to
Anga, from the Himalaya in the North to Vidarbha, Andhra in the South, and includ¬
ing the South-Eastern tribes” (Witzel 1989, 244). There does seem to be significant evi¬
dence to accept a movement across the subcontinent from the Punjab that cannot sim¬
ply be brushed aside.

Curzon (1855), however, who, as I have noted, believed that the Indo-European home¬
land was in India, was not prepared to allow anything more than just this:

Is it legitimate ... to infer that because the Aryans early spread to the South . . . and
extended themselves over the peninsula, they also originally invaded, from some unknown
region and conquered India itself? If so, the same argument might be applied to the origin
and spread of the Romans, who might be presumed to have invaded Italy from some
external unknown region, because they early spread their conquests to the south. . . . But
we know from authentic history that the Romans arose from one city and region in Italy:
that . . . they gradually extended themselves over and subjugated those territories which
subsequently formed one vast empire. (189)

Despite such objections, Muir’s interpretation prevailed and is still a prominent sup¬
port of the Aryan migration position: “The known historical expansion of Indie from
north to both the east and the south, gives us every reason to deny the Indo-Aryans a
prior home in those regions” (Mallory 1989, 44, my italics). 7 Needless to say, Indig¬
enous Aryanists see things in the same vein as Curzon: “We may notice a greater ac¬
quaintance with Central and Eastern India in the latter [texts], showing perhaps the
shift of the seat of Vedic Civilization more inland. But such a shift would be a matter
of internal history and could have no bearing on the question of the Rigvedics hailing
in 1500 b.c. from beyond the Afghanistan-Punjab complex” (Sethna 1992, 14).

Just as this book is going to press, a new publication (Talageri, 2000) attempts to
undermine these notions of a west-to-east spread of the Indo-Aryans in the subconti¬
nent based on the geographical parameters of the texts. Restricting his focus to the ten
mandalas ‘books’ of the Rgveda, Talgeri establishes an internal chronology of this text
consisting of four periods: Early (mandala 6, 3, and 7, in that order); Middle ( mandala
4 and 2, in that order); Late ( mandala 5, 8, and 9, in that order), and Very Late (mandala
10). He considers Mandala 1 to cover a period from pre-Middle (but post-Early) to the
Very Late period. Talageri’s method involves establishing a relative chronology of the
various composers of the hymns. If, for example, mandala A is composed by someone
who is an ancestor of the composer of mandala B, or if the composer of mandala A is
considered, in mandala B, to be a figure in the past, then Talageri assumes mandala A
is older. He also considers the lineages of kings and rsis such that if mandala C refers
to a contemporary king or rsi and mandala D refers to this same rsi or king as a figure
in the past, he assumes mandala D to be older.

He feels his ordering ( mandala 6, 3, 7, 4, 2, 5, 8, 9, and 10) is confirmed by the fact
that the oldest mandala demonstrate the most rigid family structure—every verse in mandala
6 is composed by members of one branch of one family of composers—while at the
other end of the spectrum, the hymns, mandala 10 , being composed by rsis from almost
every family, have the loosest family structure and contain a large number of hymns of
unknown composition. He also notes that the hymns in the older mandalas of his schema
which are composed by the descendants of an important or eponymous composer, are
generally attributed to that ancestral composer, while in later mandalas, the hymns are

66 The Quest for the Origin of Veiiic Culture

attributed to the actual composer himself. He takes this development as further confir¬
mation of his ordering.

The final step of this method is to examine the geographical horizons expressed in
this chronological sequence. In mandala 6, the Aryans were settled in the region to the
east of the SarasvatT, viz, in present-day Haryana and Uttar Pradesh. Toward the end of
the Early Period, mandalas 3 and 7, they had expanded westward into the Punjab, and
by the Middle and Late Periods, the geographical horizons had spread westward as far
as the southeastern parts of Afghanistan. In short, “the evidence of the Rigveda is so
clear that it brooks no other conclusion except drat the Vedic Aryans expanded from
the interior of India to the west and northwest” (123-24). 8

While this version of dtings awaits a response from Witzel, we should outline the
picture produced from the latter’s reading of the same texts. In terms of the internal
ordering of the texts, Witzel (1995b) follows Oldenberg who as early as the nineteenth
century had noted that the oldest mandalas, 2-7, are the collections of the clans of poets.
Within this core, the mandalas are arranged according to the increasing number of hymns
per book. Within each mandala, in turn, the hymns are ordered according to deity,
with Agni placed first, followed by Indra, followed by the other gods. Then, within the
sections for each deity, the hymns are arranged according to decreasing number of stan¬
zas per hymn. In case the number of stanzas in particular hymns are equal, the hymn
with more syllables is placed first. In dais way, a hymn can be immediately found by its
family, deity, and meter.

Of course, this internal ordering need not correspond to chronology of composition
(although it can be used to identify late additions if they disrupt this pattern). Witzel
notes that “all we can say with confidence is that book 10, as such, is late but judgment
must be exercised for each individual hymns. Some in book 8, sometimes even in book
1 and 10, can be as early as the ‘family books’” (1995b, 310). In order to fine-tune the
chronology of the hymns within the Rgveda, Witzel also accepts the internal relation¬
ship between the poets and kings to be decisive, and couples these individuals and the
tribes with which they are connected with the geographical horizons of the texts that
featute them; however, he produces a rather different picture from Talageri’s.

Witzel (1995b) focuses his attention on the pahca jana, the five tribes of the Rgveda.
He finds that four of them, the Yadu-Turvasa and the Anu-Druhya, are regarded as
already settled in the Punjab at the time of the composition of the Vedic hymns with
“only the dimmest recollection of their move into South Asia” (339). The Purus are the
next to arrive on the scene, “although their movement into the subcontinent had also
become a done deed by the time most of the Vedic hymns were composed” (339). They
dominate the remainder of the “five people.” Meanwhile, Witzel sees a subsection of
the Purus, the Bharatas, situated in Afghanistan, on the grounds that their leader,
Divodasa, is fighting his enemy Sambara, who “was probably an aboriginal tribal chief
in the mountainous Borderline zone” and is said to possess hill fortresses (332).

After the defeat of Sambara, Witzel finds the Bharatas intruding into the subconti¬
nent where they defeat the Purus in the famous battle between the Bharata chief Sudas,
a descendant of Divodasa, and the ten kings: “The entire book seven is a snapshot of
history: the incursion of the Bharata into the Panjab from across the Sindhu, and their
battle with the “Five Peoples’ and the Puru” (1 995b, 337). He sees the geographical
referents of this battle hymn shift from west to east, beginning with the Bharata priest

Vedic Philology 67

Vasistha’s crossing of the Sindhu river, followed by the actual battle of the ten kings on
the Parusni river (modern-day Ravi), and culminating with the Bharatas eventual arrival
on the Yamuna river (335). Emerging victorious, the Bharatas then settle on the SarasvatT
(which would appear to be a movement back to the west), and this area becomes “the
heartland of South Asia,” considered in Rgveda 3.53.11 to be the center of the world
(339). In short, where Indigenists would simply see strategic maneuverings backward
and forward in battles between rival clans native to the North of the subcontinent itself,
Witzel finds “successive waves of migrations,” resulting in frequent warfare and shift¬
ing alliances (337).

Conclusion

The Rgvedic texts were read in the political context of nineteenth-century philology, which
has been outlined in chapter 1. This certainly influenced the choice of possible inter¬
pretations placed on such words as anasa and on the battles of the Aryas and the Dasas.
The racial interpretations extrapolated from the texts to support an Aryan migration
have been justly challenged by both Indian and, albeit after the lag of a century, West¬
ern scholars. Their place in serious discussions of the Indo-Aryan problem is highly
questionnable.

Geographically, however, the textual evidence for immigration is more persuasive.
The sequence of texts does seem to suggest a movement of the Brahmanic geographical
horizons from the Northwest to other parts of India. Nonetheless, the Indigenous re¬
sponse needs to be considered: the texts give no obvious indication of a movement into
India itself. Indigenous Aryanists, on the whole, are prepared to accept a shift of popu¬
lation from the SarasvatT region eastward toward the Gangetic plain (with increasing
contacts with the South) in the historical period. (As we shall see in chapter 7, this is
sometimes correlated with the drying up of the SarasvatT and the eastern drift of dte
Indus Valley sites from the Mature to the Late and post-Harappan period.) But they do
not feel compelled to then project this into preconceived hypothetical movements into
the subcontinent itself in the pre- and protohistoric period.

It seems reasonable to accept that the Indo-Aryans spread into the east and south of
the subcontinent from the Northwest. But to what extent can it be legitimately argued
that they might have been indigenous to the Northwest itself? At this point, other dis¬
ciplines must be introduced into the debate, and so it is to these that I turn in the
ensuing chapters. But first, it is important to outline the history of Sanskrit in the field
of Indo-European studies over the last two centuries. As we have seen, the demotion of
India as the favored Indo-European homeland in the early nineteenth century had a lot
to do with the demotion of Sanskrit from its status as the original tongue of the Indo-
Europeans to a more secondary and reduced role as a daughter language that might be
even younger than some of its siblings.

Indo-European Comparative Linguistics

The Dethronement of Sanskrit

Since no remnant of the original Proto-Indo-European language has been discovered,
Indo-European historical reconstruction consists of comparing cognate words and gram¬
matical forms in the various historical Indo-European languages and deducing the original
forms, or protoforms therefrom. Obviously, such an endeavor is deductive, since, by
definition, there are no written documents in a protolanguage so it can never be frilly
verified. Let us take a moment to note, along with Pulgram, that this protolanguage
does not correspond to any de facto reality but to “something of a fiction,” or, in Zimmer’s
terms, (1990a), not to “one of the languages spoken in an unknown antiquity by uni¬
dentified people but as a reference tool in discussing the history and development of
the different Indo-European languages” (313). Pulgram (1959) further warns that “we
must not make the mistake of confusing our methods, and the results flowing from
them, with the facts; we must not delude ourselves into believing that our retrogressive
method of reconstruction matches, step by step, the real progression of linguistic his¬
tory.” He continues: “We now find ourselves in possession of two entirely different
items, both of which we call Proto-Indo-European: one, a set of reconstructed formulae
not representative of any reality; the other, an undiscovered (possibly undiscoverable)
language of whose reality we may be certain.” The difference between the two should
always be kept in mind: “Arguing about ‘Proto-Indo-European’ can be meaningful and
fruitful ... if we always explain whether we are talking about the one or the other—
which, as we well know, we do not do” (424). In short, we know there was a Proto-
Indo-European language; we do not know to what extent our reconstructions approxi¬
mate it.

The Indo-Europeans are purely a linguistic entity: it was comparative Indo-European
linguistics that necessitated the existence of a group of people whose language evolved

Indo-European Comparative Linguistics 69

into the classical and modern languages of Europe, Iran and India, and that co-opted
archaeology, anthropology, and other disciplines into a quest for empirical evidence of
their physical presence. Historical Indo-European linguistics got its start only when the
Paninian analysis of an Indo-European language, Sanskrit, became known in Europe;
the field was completely dependent on a mastery of Sanskrit grammar which still plays
an indispensable role in Indo-European studies. Despite being naturally advantaged can¬
didates for such studies, Indian scholars have made a negligible contribution to this
field. While there have been a few outstanding Indian historical linguists—trained, for
the most part, in either Europe, Calcutta or Poona—their number has been insignifi¬
cant; with the exception of a history of Indo-European studies in Japan, the field of
Indo-European historical linguistics has been almost exclusively a Western (and, most
particularly, at least in its earlier phases, a German) domain.

One explanation for this regrettable dearth of Indian participation was made by the
linguist D. D. Mahulkar (1992) in an attempt to promote die case for the study of his¬
torical linguistics in India:

It was a matter of great pride for us to know that the European scholar’s search for the
“origin of language” had found its first springs on the Indian soil in the classical language
of India, Sanskrit. Sanskrit since then began to play a role altogether undreamt of by its
traditional grammarians. ... [It became) the most authentic tool in the hands of European
scholars for exploration into the origin of their languages. . . . but doubts started creeping
into the science after the law of palatals (1870) came to be formulated . . . [which was] to
dethrone Sanskrit from its pedestal of being the dialect nearest to the original language. . . .
it appears that Indian linguists of the time gave up the study of linguistics out of the dis¬
illusionment over the see-saw position of Sanskrit in what Whitney named as “European
Sanskrit Science.” There was an overtone of emotionalism in their action. (4-6)

As discussed previously, Sanskrit was initially considered by early scholars such as Schlegel
and Eichhoff to be the original mother language or, at least, as suggested by scholars
such as Bopp and Pott, almost identical to it despite containing some innovations. Ini¬
tially, the homeland was to a great extent situated in India because it seemed logical to
locate it where the original, or oldest, languages were spoken. The complete dethrone¬
ment of Sanskrit to the point where it was no longer deferred to as being the preserver
of the most archaic linguistic forms in most circumstances took the best part of a cen¬
tury of comparative philological research. 1

The Law of Palatals and the Discovery of Hittite

The law of palatals was the primary linguistic formula arrived at by the comparative
method that dramatically shattered Sanskrit’s preeminent status as the most venerable
elder in this reconstructed family. It was preceded, in this regard, by an influential article
by Grassman published in 1863. According to what came to be known as Grimm’s
law, certain regular changes had taken place in the Germanic languages when compared
to other Indo-European languages such as Sanskrit. These included the observation that
voiced stops had become voiceless stops (e.g. b had become p) in Germanic. 2 However,
an irregularity to this sound change was evident. Sanskrit bandh - corresponded to Gothic
bindan, where Grimm’s law would have predicted p instead of b in tire Gothic. Assum-

70 The Quest for the Origin of Vedic Culture

ing Sanskrit to be the more pristine language, scholars intially thought the irregularities
in correspondences between Germanic and Sanskrit were due to aberrations in Ger¬
manic. However, Grassman determined that this “irregularity” was due to Sanskrit, not
Germanic, and was the result of a separate sound change. His observations resulted in
a law which states that an aspirated sound looses its aspiration when followed by an¬
other aspirated sound. So an original Proto-Indo-European form *bhendh- had become
bandh- in Sanskrit. The b in band h-, then, was due to innovations in Sanskrit, which up
till then was generally assumed to have preserved the oldest forms, and not due to in¬
novations in German as had originally been assumed. Sanskrit’s deviation in this re¬
gard resulted in the reconstructed (i.e., unattested in any language) Indo-European forms
being denoted with a preceding asterisk—as in *bhendh. Previously, the Sanskrit forms
had generally been accepted as representative of the proto-language.

The law of palatals solidified this demotion more conspiciously and sealed Sanskrit’s
fate as a sister language to Greek, Latin, Germanic, and the other Indo-European lan¬
guages, rather than the more-or-less exact preserver of the original mother tongue.
The final formulation of the law of palatals was the result of a series of discoveries by
a number of European scholars. 3 Put briefly, Sanskrit has three primary vowels: a, i,
and u. Up until the 1870s, scholars believed that this vowel triad represented the
original Indo-European vowel system. However, while the occurrences of the i and u
vowels in words cognate to Sanskrit and the other Indo-European languages was
unproblematic (i.e., where a Sanskrit word contained an i or u, its cognates did like¬
wise), the a vowel proved to be unpredictable. In words where Sanskrit had preserved
an a, other Indo-European languages revealed cognate forms with either a, e, or o
with no apparent consistency. 4 The method of historical linguistics is based on the
tenet that all sound changes are rule-governed, but there did not seem to be any dis¬
cernible rules governing these particular vowel correspondences. Although linguists
found this multiplicity of Indo-European vowels in cognate forms bewildering and
seemingly erratic, they assumed that Sanskrit had best preserved the original IE pho¬
nological system and that therefore Sanskrit a was the protosound that the other lan¬
guages somehow had corrupted into a, e, and o. But they could not determine the
phonemic laws governing such changes.

The law of palatals reversed all this by postulating that it was the other languages
such as Greek and Latin that, in addition to i and u, had also faithfully preserved the
original vowels a, e, and o from an original Indo-European pentad i, u, a, e, and o:
Sanskrit had been the innovator by merging the latter three vowels into a single a to
form its vowel triad. 5 Specifically, the law postulated that velars are replaced by palatals
in the environments where Sanskrit a corresponds to e in other dialects, and that there¬
fore Sanskrit once also contained an e in those environments The law was perceived
most clearly in the reduplicated perfective form cakdra ‘he/she did’. The basic charac¬
teristic of this preterit tense, termed lit by Paninian grammarians, is that the first syl¬
lable of the root reduplicates. However, guttural (velar) stops are irregular in this re¬
gard, being replaced by palatals in the reduplicated form. Thus, kr yields cakdra, and
gam- yields jagdma. Within Sanskrit, there is no explanation for this phenomenon—the
reduplication should logically have resulted in kakdra and g agdma. Greek, however,
showed that the reduplicated syllable has an e vowel. This was the clue that alerted the
philologists.

Indo-European Comparative Linguistics 71

Scholars argued that according to the law of palatals, the reduplicated form of kr
was originally kekora, the first k being a reduplication of dre root syllable in accordance
with the regular laws of reduplication. However, it was proposed that due to the influ¬
ence of the postulated vowel e following the reduplicated k, this latter transformed into
c. The law of palatalization claimed that the vowel following the reduplicated syllable
could not have been an a; otherwise, there would have been no stimulus for the pala¬
talization of the guttural to have occurred in the first place. The change seems natural,
in contrast, if the reduplicated vowel was an e. The reason is simply one of euphonic
ease of articulation. Front vowels, such as e, cause palatalization, whereas back vowels,
such as a, have no such effect. As Pedersen (1931) puts it: “This Indian law could not
be seen unless one dared assume that the vowel system in Sanskrit is quite the reverse
of primitive, and that the Latin e in que ‘and’ (Greek te) is older than the a in Sanskrit
ca ‘and’” (280). The thought was revolutionary. In one stroke, there was no more mys¬
tery concerning the vowel correspondences: method had finally been brought to the
madness. The hitherto perceived conservatism of Sanskrit, however, was the casualty of
this advancement in linguistics; it had now officially lost its exclusive claim of being the
most archaic member of the family in all respects.

A later hypothesis that further significantly demoted Sanskrit from its venerable po¬
sition in the eyes of many linguists as the most intact member of the Indo-European
language family, was the discovery of the laryngeals in Hittite. Saussure initially postu¬
lated the existence of certain unknown resonant sounds in the Proto-Indo-European
language. 6 These resonants, which Saussure called coefficients sonantiques (and repre¬
sented as A and O) but were later called laryngeals, disappeared in the daughter Indo-
European languages, but their original existence was deduced by Saussure from the effects
they had left behind on neighboring vowels before becoming extinct. 7

Over half a century later, in 1927, Kuryiowicz examined the Hittite documents that
had been discovered in Anatolia during the First World War in the light of Saussure’s
contentions. He found that they actually contained a h phoneme in the exact linguistic
locations that Saussure had predicted that his A resonant should have once existed.
This h phoneme was termed a laryngeal by analogy with Semitic laryngeals, which were
also lost after they had affected the vowel in a fashion similar to Saussure’s proposed
coefficients. What had been postulated by the inferential process of comparative linguis¬
tics had now been demonstrated as factual by empirical, inscriptional data. 8 While the
discovery of die laryngeal greatly validated the efficacy and validity of the process of
historical reconstruction, the discovery of Hittite and, to a much lesser extent, Tocharian
(which will be discussed later) has significantly altered the appearance of the reconstructed
protolanguage not only in phonology but also in morphology and lexicology. As an
aside, as has been pointed out, these discoveries underscore the tentative nature of Proto-
Indo-European reconstruction: any further future uncovering of presently unknown lan¬
guages may cause linguists to alter their present reconstructions and produce a signifi¬
cantly altered reconstructed language that is different from the one linguists have
reconstructed today. As Pulgram (1959) points out, “when a new dialect is discovered,
the inventory of Proto-Indo-European phonemes may have to be revised, as indeed it
was after we learned about Tocharish and Hittite. Needless to say, the inventory of real
Proto-Indo-European is not subject to revision on the evidence of modern discoveries”

(424).

72 The Quest for the Origin of Vedic Culture

Be that as it may, for our purposes, the discovery of the laryngeals, although still the
subject of debate among linguists, reveals a written Indo-European language that had
preserved linguistic features that were more archaic than Vedic. This “demotion” of
Sanskrit has been ongoing ever since; although Indo-European studies is still heavily
dependent on Sanskrit for any attempt at reconstruction (indeed, there is no way of
determining if there even would have been a discipline of Indo-European studies without
Sanskrit), it has been claimed that every commonly accepted adjustment to the recon¬
struction of Proto-Indo-European in the last century and a half has involved a further
de-Sanskritisation of the protolanguage. As in any discipline, however, there are still
“conservative” linguists who remain unconvinced about some of the new proposals con¬
cerning the protolanguage and who still accept Sanskrit as the most faithful preserver of
most of the protoforms.

Objections from India

One of the very few Indian linguists who has contested some aspects of this de-
sanskritization of Proto-Indo-European is Satya Swarup Misra (1977, 1994), who has
particularly opposed the laryngeal theory for three decades. 9 His opposition, it should
be noted, has nothing to do with the Aryan invasion debate or with issues connected
with Vedic antiquity, since at the time he first published his thesis, he accepted the
theory of Aryan invasions and the commonly accepted date of 1200 b.c.e. for the
Rgveda. 10 Misra has repeatedly argued that the theory rests on limited evidence pro¬
vided by only one member of the Indo-European language family—die Anatolian group. 11
He prefers to consider that the Hittite h symbol either had merely graphic status with
no phonemic value or was a phoneme borrowed from another language family, per¬
haps Semitic, in a manner analogous to the supposed borrowing of retroflexes into Indo-
Aryan from Dravidian (which are phonemes unattested in other Indo-European lan¬
guages and therefore not ascribed to Proto-European). 12

Nonetheless, apart from the laryngeals, Hittite is considered to have preserved other
archaic Indo-European linguistic features that Sanskrit and other Indo-European lan¬
guages have lost. Some linguists, for example, believe that it has retained the original
common Indo-European gender where Vedic, and other languages, innovated the mas¬
culine-feminine gender distinction. 13 Although other linguists reject this idea, the gap
between Sanskrit and Proto-Indo-European has progressively widened in the eyes of some
linguists. Several scholars such as Polome have argued that some features in Indie and
Greek that have always been considered archaic, such as the reduplicated perfect, are
actually late Indo-European innovations: “We must consider Greek and Indo-Aryan as
reflecting the most recent developments in the Indo-European speech community at the
time of its disintegration” (Polome 1985, 682). Lehmann (1993) acknowledges that “many
Indo-Europeanists still assume that the subgroup including Sanskrit and Greek is clos¬
est to the proto-language” but feels this position “requires extraordinary effort to ac¬
count for the morphological structure of Hittite, especially because of its similarity with
that of Germanic. It also overlooks innovations in Greek and Sanskrit that have not
been extended throughout all categories in the early stages of these languages, such as
the augment” (261). Indeed, Lehmann goes so far as to say that “the history of recon-

Indo-European Comparative Linguistics 73

structing Proto-Indo-European might be characterized as a continuing effort to liberate
it from the heavy hand of Sanskrit” (261). Of course, the last word has yet to be said on
such matters, and Polome (1985) makes a point of adding that “all of this does not
imply . . . that Indie does not show archaic features” (682). But Proto-Indo-European
appears far less Sanskritic nowadays than it was in the fable Schleicher attempted to
reconstruct in Proto-Indo-European in 1871, which was almost pure Sanskrit. 14

More recendy, since Misra (1992) has begun to reject the Aryan invasion theory on
linguistic grounds, he has become more and more suspicious of the reconstructed proto-
language. He has even reexamined the primacy of Greek and Latin a, e, and o over
Sanskrit a. He suggests that the Gypsy languages provide evidence that an original San¬
skrit a vowel does demonstrably evolve into a, e, and o in later languages. 15 The Gyp¬
sies are generally accepted as having migrated from India some time in the common era
and speak languages that are Indo-Aryan in character. The logic here is that if, in the
historic period, Indo-Aryan speaking tribes have left India and migrated to Europe, where
they are found speaking dialects that have transformed an original Sanskrit a vowel into
a, e and o, dten the same could be postulated for the protohistoric period: Indo-Euro¬
pean tribes could have left a homeland in India carrying a Proto-Indo-European a that
evolved into the a, e, and o of the later Greek and Latin languages. 16 However, Misra’s
proposal, apart from anything else, gives only a passing reference to the law of palatal¬
ization which, as has been outlined, initially caused linguists to postulate a Proto-Indo-
European a, e, and o triad, as opposed to a proto a form, in the first place.

It should go without saying that no serious linguist is about to reverse a century and
a half of linguistic research to contemplate the proposal that proto-Indo-European was
identical with Vedic, as some members of the Indigenous Aryan school might hope. 17
Of course, even allowing the more complete preservation of the Indo-European vowels
in, say, Greek or other European languages, and of the laryngeals in Hittite, or of the
archaic features in Lithuanian, Indian scholars can still argue that Vedic is, overall, the
language that has most completely retained the Proto-Indo-European character despite
the present trend among many linguists to consider the distinctive Vedic features to
have been later innovations. But language is never static. Linguists cannot be expected
to accept Vedic as an immutable linguistic entity that somehow transcends the transfor¬
mations visible in every language known to man (transformations that are, indeed, per¬
ceivable in the Vedic texts themselves). The discourse of an eternal, unchanging Veda
is a legitimate one for the ashram, not the academy. 18

Conclusion

Whatever might have been the real nature of Proto-Indo-European, for the purposes of
this study, the existence of various linguistic stages of Indo-European more archaic than
Vedic is irrelevant (or, at least, peripheral) to the problem of the origin of the Indo-
Aryans. In the absence of compelling counterarguments, there are no grounds to ques¬
tion that Proto-Indo-European resembles, more or less, the reconstructed entity diligently
assembled to the satisfaction of most historical linguistics with its e and o vowels, its
laryngeals, and the rest of its carefully formulated components. Unless this linguistic
evidence is deconstructed in a thoroughly comprehensive methodological way, there are

74 The Quest for the Origin of Vedic Culture

no a priori reasons to reject the basics of the Proto-Indo-European language as accepted
by most linguists. 19 The real problem ultimately facing Indigenous Aryanists boils down
to whether, or more specifically how, this Indo-European language could have origi¬
nated in some area of India. If this proposal cannot be argued convincingly, then it at
least must be demonstrated whether and how India can avoid being excluded as one
potential homeland among others.

It is all well and good to insist that there is no evidence supporting the theory of an
external origin of the Indo-Aryan language, but how are the connections between Vedic
Sanskrit and the other members of the Indo-European language family to be explained?
That cognate languages must, by definition, have a common linguistic and geographic
origin is an assumption that few have challenged. 20 The history of the Indo-Aryan group
of languages must always be reconstructed from within the context of its membership
in the greater Indo-European language family of which it is an unassailable member.
Any theories pertaining to one member of the family must be in harmony with the data
connected widr the other members of the family. If Indo-Aryan is to be considered in¬
digenous to the subcontinent, then how is its relationship to the other Indo-European
languages to be explained? Either the Indo-Aryan languages entered the subcontinent
from an external geographic origin or, if it is to be argued that this group is indigenous,
at least to the Northwest, the unavoidable corollary is that northwest India is the geo¬
graphic origin of all the other Indo-European languages, which must have emigrated to
the West from there. Indigenous Aryanists have to confront and address this language
connection, with its inescapable requirement of a common language origin.

However, many scholars tend to avoid pursuing their claims of Aryan autochthony to
this inevitable conclusion. Indeed, with a very few exceptions, most Indigenous Aryanists
tend to completely ignore the linguistic evidence altogether. Others are almost cavalier in
their dismissal of it: “All the comparative philological speculations, . . . according to some
of us, provide the greatest stumbling-block to the appreciation of a multi-linear archaeo¬
logical development of culture and civilisation in different parts of India” (Chakrabarti
1995, 429); “Ancient Indian history is ripe for a thorough revision. . . . one can begin by
clearing away the cobwebs cast by questionable linguistic theories,. . . using every avail¬
able modern tool from archaeology to computer science” (Rajaram 1995, 230).

There seem to be two very basic reasons for this neglect. Seeing the morass of oppos¬
ing opinions supposedly based on the same linguistic evidence, some scholars have
rejected the whole Indo-European enterprise as hopelessly speculative and inconclusive.
To put it another way, they are not interested—this is primarily a European preoccupa¬
tion. 21 Alternatively, for those interested in Indo-European historical linguistics, they
soon find that the field is inaccessible to the dilettante and requires organized study that
is not easy to come by in India. Current scholarly publications in Indo-European pre¬
suppose acquaintance with almost two cenmries of linguistic discoveries (much of it in
German and French and, more recently, Russian), as well as with at least a number of
principle Indo-European languages (including, at least, Sanskrit, Greek, Latin, Hittite,
and Germanic). 22 As mentioned earlier, although there have been, and still are, some
exceptional historical linguists in India, their numbers are few. This is not a subject
that is fruitfully pursued, and it has always remained predominantly a Western field of
scholarship. It should be noted that there is very little facility in India for the study of

Indo-European Comparative Linguistics 75

historical linguistics. When I conducted my research (1994-95), only two universites in
the subcontinent were actively offering adequate courses in this field. 23

Perhaps the dethronement of Sanskrit as the most archaic Indo-European language
did initially contribute to a loss of interest in this field as suggested by Mahulkar, but
one suspects that present-day concerns are more pragmatic, at least in terms of the aca¬
demic powers that be. As is increasingly apparent even in areas of the Western acad¬
emy, sociolinguistics or modern language study produce more tangible fruits for dwindling
government funding. Indian scholars, for the most part, simply do not have the train¬
ing to embark on a scholarly critique or evaluation of the vast specialized field of Indo-
European studies. For the most part, with some exceptions, Indigenists wishing to tackle
the Indo-Aryan issue are left with the option of either ignoring the linguistic dimension
of the problem or attempting to tackle it with sometimes hopelessly inadequate qualifi¬
cations. Typically, the formulations of the Indo-Europeanists, being inaccessible, are
dismissed in frustration as highly speculative and irrelevant. This attitude and neglect
significantly minimizes the value of most Indigenist publications. 24

In any event, as we shall see, a few scholars have argued that the linguistic evidence
could just as well be reconfigured to postulate that even India might have been the Indo-
European homeland. This challenge, when made by the more sober members of the
Indigenous Aryan school, does not so much aim to prove that India factually was dte
original homeland as to assert that the linguistic evidence is not sufficiently conclusive
to fully determine where the homeland was. In view of this challenge, we must, as a
purely theoretical exercise, consider whether the linguistic evidence can exclude the
possibility of a South Asian homeland as one candidate among others.

Linguistic Substrata in Sanskrit Texts

Another principal reason that South Asia had been excluded relatively early as a poten¬
tial Indo-European homeland was that it showed evidence of a pre-Indo-European lin¬
guistic substratum—considered to have been of Dravidian, Munda, or other unknown
languages. I use the term substratum in the sense of an indigenous language being sub¬
sumed and displaced by an alien incoming language. In this process, die indigenous
language affects the dominant language by depositing into it its own linguistic features,
such as vocabulary or morphology. These features form a substratum in die dominant
language that can be discerned by diligent linguists. Such evidence is solid reason to
exclude India and must detain us at length. Southern Europe had also been eliminated
for the same reasons—northern Spain and southwestern France had a pre-Indo-Euro¬
pean substratum in the form of Basque, and Italy had Etruscan. This method, which
has sometimes been called the exclusion principle, initially serves to at least delimit the
range of candidates for the Proto-Indo-European homeland from within the massive
area where the cognate languages were and are spoken. Dravidian, Munda, Basque,
and Etruscan are not Indo-European languages, suggesting that Indo-European was not
the original language family in those areas and therefore must have intruded from else¬
where. The evidence of such a substratum remains one of the most cited and compel¬
ling reasons for accepting the verdict that the Indo-Aryan language must have had an
origin external to the Indian subcontinent. As Emeneau (1954) puts it: “This is our
linguistic doctrine which has been held now for more than a century and a half. There
seems to be no reason to distrust the arguments for it despite the traditional Hindu
ignorance of any such invasion, their doctrine that Sanskrit is the ‘language of the Gods,’
and the somewhat chauvinistic clinging to the old tradition even today by some Indian
scholars” (282).

Linguistic Substrata in Sanskrit Texts 77

The best-known study of the Dravidian languages was made by the Reverend Robert
Caldwell in his pioneering Comparative Grammar of the Dravidian Languages in 1856
(although, in actuality, the first publication on the Dravidian language family was by
Francis Whyte Ellis, as far back as 1816—the year of Bopp’s famous comparative gram¬
mar). 1 Until the founding of the Madras Orientalist school, scholars, who had assumed
that all the languages of the subcontinent had a common origin, had been almost exclu¬
sively concerned with the Indo-Aryan languages because of the relationship these lan¬
guages had with the classical languages of Europe and the exciting implications this
connection suggested for the origins of Western civilization. In 1849, Alexander Campbell
continued the work of Ellis and was able to demonstrate that “it has been very generally
asserted, and indeed believed, that the Teloogoo has its origin in the language of the
Vedums. ... my inquiries have led me to the opposite conclusion.. . . Teloogoo abounds
with Sanskrit words. . . . nevertheless, there is reason to believe that the origin of the
two languages is altogether distinct” (xv-xvi). The breakthrough for these scholars was
originally triggered by the extensive analyses of the traditional grammarians of India,
who had noted the distinction between the Sanskrit words and the non-Sanskritic desya
ones. This alerted these linguists to the possibility of non-Indo-Aryan languages in the
subcontinent.

The discovery of non-Sanskritic languages in the South provided important material
that seemed to militate against the idea of the Indo-Aryans as indigenous to India—or at
least to any of it except the Northwest. In a series of articles in the early 1840s, Stevenson
(1844) laid out the implicatioris of this non-Sanskritic, indigenous language base, which
he perceived even in the northern vernaculars:

If we can trace a language wholly different from the Sanscrit in all the modern dialects,

... it will seem to follow, that the whole region previous to the arrival of the Brahmans
was peopled by the members of one great family of a different origin. ... I call the Brah¬
mans a foreign tribe in accordance with indications derivable from the cast of their fea¬
tures, and the colour of their skin, as well as their possessing a language which none of
the natives of India but themselves can even so much as pronounce; and the constant
current of their own traditions making them foreign to the whole of India, except per¬
haps a small district to the north-west upon the Ganges. (104)

In another article, Stevenson (1851) elaborated more fully on his theory, which, is more
or less an account that can still be found in texts on Indian prehistory. He envisioned
tribes entering the main part of India, defeating the aborigines by force of arms, expel¬
ling them from the northern regions, and enslaving those that remained (73). By the
middle of the nineteenth century, most Western scholars had accepted the existence of
the Dravidian languages as conclusive proof that the Aryans could not have been indig¬
enous to the subcontinent: “That the Arian population of India descended into it about
3000 years ago from the north-west, as conquerors, and that they completely subdued
all the open and cultivated parts of Hindustan, Bengal and the most adjacent tracts of
the Deccan but failed to extend their effective sway and colonization further south, are
quasi historical deductions confirmed daily more and more” (Hodgson 1848, 551).

To this day, a few linguistic islands of Dravidian languages, of which Brahui occu¬
pies the largest area, exist in the North of India surrounded by an ocean of Indo-Aryan
languages. They are generally explained as being isolated remnants that had somehow
avoided being engulfed by the incoming tides of Indo-Aryan speakers encroaching on

78 The Quest for the Origin of Vedic Culture

what had originally been a Dravidian language area spread throughout much of India.
Moreover, a few words in the Rgveda, progressively more words in later Vedic texts,
and a much larger number of words in later Epic texts were identified as being loanwords
principally from Dravidian but with some forms traceable to Munda. The borrowings
were held to have ceased shortly after the Epic period. This seemed to fit neatly into the
hypothesis that when the Vedas were composed, the Indo-Aryans had only recently arrived
into Dravidian-speaking India and had absorbed only a few words from the local
Dravidian substratum. 2 As time went by and the Aryan speakers merged completely
with the native population, more and more Dravidian words were absorbed, which
emerged increasingly in the later texts. Sometime before the common era, when the
Dravidian languages in the North had been completely subsumed by Aryan dialects,
the borrowings ceased. This whole process is usually understood as being the result of
the Aryans themselves adopting new items of the local lexicon and/or of bilingual
Dravidian speakers increasingly adopting the new Aryan languages until they eventu¬
ally became completely co-opted into speaking Indo-Aryan, but not without preserving
a significant number of words and other linguistic features from their own languages. I
will deal first with these other, nonlexical items. Here, as in the remaining chapters on
Linguistics, I will try to keep technical details to a minimum.

Linguistic Innovations in Sanskrit

Vocabulary was not the only Dravidian element that seemed to have surfaced in Indo-
Aryan: a number of syntactical and morphological features common to Dravidian, but
alien to other Indo-European languages, had been identified since Caldwell’s work in
1856. None of these features were explicitly present in other Indo-European languages,
but they did occur in Sanskrit, Dravidian, and other South Asian languages. Such shared
syntactical and morphological features reinforced the hypothesis drat the Indo-Aryans
were intruders to the subcontinent and that bilingual Dravidian speakers who preceded
them in the North of India had welded some of their own local syntactical features onto
the encroaching Aryan languages they were adopting. This idea of syntactical conver¬
gence has developed into the established paradigm of South Asia as a linguistic area. 3
In the case of Sanskrit, these syntactical innovations were generally held by most schol¬
ars to be due to a local substratum of Dravidian, which triggered this linguistic subver¬
sion (most recently Emeneau 1980; Kuiper 1991). Munda is another candidate (Witzel
1999, forthcoming b). Some scholars see a combination of both. Southworth, for ex¬
ample, on the basis of geographic dialect variations in the Asokan inscriptions as well
as in New Indo-Aryan languages, proposed that “while the population in the western
areas . . . was probably mainly Dravidian speaking, in the Gangetic plain . . . the IA
language was taken up by a predominantly Tibeto-Burmese population” (1974, 222). 4 An
unknown primordial language, which was pre-Indo-Aryan, Dravidian, or Munda, called
“language x” by Masica (1979) has also been identified in the texts.

As noted, bilingualism is held to be one of die mechanism of the innovations, with
most linguists painting a picture of non-Aryan natives from the lower strata adopting
the intruding elite language of Indo-Aryan (although Southworth 1979 sees social equals),
but adapting it somewhat to accommodate certain lexical, phonemic, morphological,

Linguistic Substrata in Sanskrit Texts 79

and syntactic features of their own languages (e.g., Kuiper 1991, 96). Although this
substratum language subsequently disappeared in North India, its original existence
there could be inferred from the linguistic clues it had left behind in Sanskrit and
later Indo-Aryan languages in the form of non-Indo-European syntax, vocabulary and
root structure.

There are, in particular, three linguistic feanires that are innovative in the Rgveda
and that have been the subject of the most discussion. Phonologically, there is the intro¬
duction of retroflexes, which alternate with dentals in Indo-Aryan; morphologically there
are the gerunds, absolutives or verbal participles (e.g., hatva instead of jaghanvan); and
syntactically there is the use of iti, a postposed quotative marker. Other features have
also been noted by linguists, 5 but since my focus is primarily on the Rgveda these items
are among those particularly relevant here and serve as suitable exemplars for the pur¬
pose of this analysis.

Caldwell ([1856] 1875) who was the first to thoroughly examine such innovations,
noticed instances where the North Indian vernacular languages shared certain particu-

80 The Quest for the Origin of Vedic Culture

lars with the Dravidian languages that were not shared with other Indo-European lan¬
guages (59). 6 Although accepting the hypothesis that the derivation of the North Indian
languages from Sanskrit had transpired not so much because of natural processes as
from the “overmastering, overmolding power of the non-Sanskritic element contained
in them,” Caldwell rejected the possibility that these non-Sanskritic elements—the sub¬
stratum, in modern parlance—could have been Dravidian: 7 “Whatever the ethnological
evidence of their identity may be supposed to exist,—when we view the question philo-
logically, and with reference to the evidence furnished by their languages alone, the
hypothesis of their identity [as Dravidian] does not appear to me to have been estab¬
lished” (58). Caldwell pointed out that the structural similarities common to Dravidian
and the Indo-Aryan languages were also shared with other non-Aryan languages;
Dravidian was not the only candidate that might have influenced the development of
Sanskrit.

Such arguments would be of merely historical interest were they not to keep resur¬
facing. In 1924 and again in 1929, Jules Bloch reiterated some of Caldwell’s objec¬
tions, insisting that “in the present state of our knowledge, it is impossible to affirm
that the form taken by the Aryan language in India is due to its adoption by a Dravidian-
speaking population. If there is a substratum at all, this substratum could just as equally
be looked for in other families, especially the Munda family” (1924-30, 20; my trans¬
lation). 8 Bloch also remarked that Brahui quite easily could have migrated from central
India to the North in the same way as the Dravidian-speaking Oraon and Maler tribes
had migrated north to Chota Nagpur (2). Likewise, the Dravidian-speaking Kurukh and
Malto tribes live in territory with Munda place-names and still maintain traditions re¬
ferring to their migration into Munda territory (Bloch 1946; see also Hock 1996). Like
Caldwell, Bloch rejected the idea that Dravidian could have been the origin of certain
phonemic and syntactic innovations in Indo-Aryan, and stressed that Dravidian was
not the only language sharing the syntactical features that distinguished Indo-Aryan from
other Indo-European languages, some of which, he argued, could have been internal
developments rather than borrowing. 8

Over the last two decades, H. H. Hock (1975, 1984, 1996) has strongly challenged
the notion that the linguistic convergences in the Indian languages could only have been
due to a Dravidian substratum. For the most part, he argues that many of the innova¬
tive features Indo-Aryan shares with Dravidian actually have parallels in other Indo-
European languages and are therefore more likely to have been internal developments
and not borrowings from any other language at all. Citing examples of retroflexion
occurring in other Indo-European languages, Hock proposed that this trait could have
been an indigenous development (a possibility Burrow also felt was perfectly valid).
Likewise, he listed occurrences of gerunds and participles (absolutives) in other Indo-
European languages, as well as usages of the quotative marker iti. Hock (1984a) argues
that the subject-object-verb (SOV) syntactical positioning Indo-Aryan shares with
Dravidian may have existed in Indo-European itself in its eastern area (97). He also
claimed that all the uses of api— ‘also’, ‘and’, ‘even’, ‘totality’, and ‘-(so)ever’—have par¬
allels in Indo-Iranian and Indo-European (103-104).

If Hock is corect that all these features could have developed in the other Indo-Euro¬
pean languages that had no contact whatsoever with Dravidian, then their occurrence
in Indo-Aryan must also be possible without requiring Dravidian influence (see Kuiper

Linguistic Substrata in Sanskrit Texts 81

1991, 10, for criticism). Like several linguists, Hock argues that retroflexion can be ex-
plained by purely internal developments; he notes that other Indo-European languages,
albeit the more recent ones of Swedish and Norwegian, have developed the feature
completely independent of any substratum. He reiterates and expands on Bloch’s (1924)
comments that the positioning and phonology of various retroflexes differ in Dravidian
and Indo'Aryan. This fact alone problematizes the candidacy of Dravidian as a substra¬
tum for Indo-Aryan and caused Tikkanen (1987, 295) to postulate that both Dravidian
and Indo-Aryan had received the feature from two separate substrata.

Hock’s model (1993) is one of language “convergence” rather than “subversion” a
process of mutual directionality or exchange between languages as the result of long¬
standing bilingual contact, as opposed to unidirectional transfers under conditions of
social inequality—a process “similar to what we find in modern South Asia” (76). He
notes a very important point: many of the early phonological differences between Indo-
Aryan and Dravidian disappear toward the modern period, but this has nothing to do
with subversion. It was caused by convergence. In other words, adstratum can account
for this convergence, not substratum. In adstratum situation, languages are geographi¬
cally adjacent to each other, as opposed to one intrusive language being superimposed
on a preexisting language.

The other type of stratum relationship, that of superstratum, entails linguistic intru¬
sions into a preexisting dominant language area by other languages that were eventually
subsumed (or that eventually proceeded on), but not before affecting the dominant lan¬
guage in some fashion. English, for example, intruded into India in the Colonial Pe¬
riod. It did not displace the local languages, but did pass many loan words into them.
Hock also argues (1993) that the whole phenomenon of shared retroflexion (which, as
noted earlier, could have been an internal development) also could have been caused by
convergence between the two languages or by mediation by a third entity (96). Hock
(1984a) concludes by encouraging scholars to look for alternative explanations to the
Dravidian substratum hypotheses, since “the claim that Dravidian influence on San¬
skrit began in pre-Rig-Vedic times must be considered not supported by sufficient evi¬
dence” (104). As we shall see, the vocabulary evidence does not favor direct contact
with either Dravidian or Munda, thus causing scholars to postulate an even earlier
unknown language. Hock (1996b) is open to the possibility that there may have been
other languages, no longer extant, that could have preceded Sanskrit. He notes, how¬
ever, that such a claim is “methodologically problematic, since it can be neither verified
nor falsified” (57).

Reluctance to accept substratum influences, and particularly Dravidian substratum,
as the cause of specific linguistic developments in Indo-Aryan has remained consistent.
The possibility of the retroflexes in Sanskrit being a spontaneous phenomenon has been
current since Buhler’s work in 1864. As Kuiper ([1967] 1974) remarks, “No agreement
has yet been reached after a discussion extending over more than a century” (138).
Deshpande (1979), in accordance with other scholars such as Emeneau, does believe
that Dravidian was the origin for this feature but argues that retroflexion did not exist
in the original Rgveda. He believes that its origin “lies in Dravidian speakers adapting
Aryan speech to their native phonology” (297). Kuiper (1991), in contrast, is quite ada¬
mant that they “must have penetrated into Indo-Aryan in a prehistoric (‘pre-Vedic’) period”
(14). Hamp (1996), while prepared to allow that “Dravidian articulation habits may

82 The Quest for the Origin of Vedic Culture

have contributed to these allophonic variations,” demonstrates that the conditions “were
laid and can be traced in the Indo-European patrimony of Sanskrit.” He laid out the
series of sound changes from purely inherited Indo-European material that arrive “by
perfectly orderly Lautgesetze' to the distinctive feature of retroflexion (721; see also Vine
1987). In other words, retroflexion can be explained purely as the result of spontane¬
ous linguistic sound processes inherent in Indo-Aryan itself: it need not be seen as the
result of a linguistic imposition from a foreign language.

Similar disagreement revolve around other such innovations in Indo-Aryan. Like Hock,
Tikkanen, for example, in his book on the Sanskrit gerund, finds that adstratum influ¬
ences could just as easily account for the commonalty of this feature between Indo-Aryan
and Dravidian. 9 Jamison, (1989) in her review of Tikkanen’s book, dismisses any talk
of substratum at all in this regard, “since one less committed to the substrate explana¬
tion can easily see mechanisms whereby the gerund could have independently acquired
the value it has when it enters history” (461). (For criticisms, see Kuiper 1991, 10).
Tikkanen’s comment about adstratum, as opposed to substratum, influence, which
parallel’s Hock’s notion of’Convergence,’ provides an alternative model tor the linguis¬
tic history of South Asia that will be further discussed later.

The Indigenous Aryan school, for the most part, seems to be either completely oblivi¬
ous to the whole linguistic substratum dimension of this issue or dismissive of it: “It
now appears that the underlying linguistic theories may be enjoying a . . . charmed life.
All theories should be validated by data independent of the theory, and not by hypothetical
constructions derived from the theory itself. Until that happens, the safest course is to re¬
gard these theories as unproven conjectures and refrain from basing any key conclu¬
sions on their claims” (Rajaram 1995, 219: italics in original). This is a conspicuous
and curious neglect. It is conspicuous because the whole theory of the external origin of
the Indo-Aryans was greatly accelerated by the “discovery” of the Dravidian languages
and of the non-Indo-European elements in Sanskrit, and this evidence needs to be ad¬
dressed by those contesting the theory. It is curious because if we ignore, for the mo¬
ment, all other types of data connected with the problem of Aryan origins, it would
seem that there might actually be some scope in all this to challenge the assumption
that there had to be a pre-Aryan linguistic substratum at all—at least in the matter of
phonology, morphology, and syntax. Several linguists, all of whom accept the external
origin of the Aryan languages on other grounds, are quite open to considering that the
various syntactical developments in Indo-Aryan could have been internal developments
rather than the result of substrate influences, or have been the result of adstratum.
Dravidian in particular, the most popular candidate, has been consistently challenged
as a possible linguistic substratum for Indo-Aryan.

Before proceeding any further, it is very' important to note that there is a serious
methodological drawback in these types of comparisons, since we cannot compare the
syntax of the Rgveda with actual contemporaneous Dravidian texts, but only with a re¬
constructed proto-Dravidian. The oldest Dravidian literary evidence that we know of
comes from epigraphs of old Tamil written in a form of Brahml script sometime be¬
tween the third and the first centuries b.c.e. The first completely intelligible, datable,
and sufficiently long and complete epigraphs that might be of some use in linguistic
comparison are die Tamil inscriptions of the Pallava dynasty of about 550 c.e. (Zvelebil
1990), two entire millennia after the commonly accepted date for the Rgveda. Postulat-

Linguistic Substrata in Sanskrit Texts

ing syntactical influence on Vedic from Dravidian material attested after such a vast
amount of time has obvious methodological limitations.

The Munda languages have left no significant literary tradition at all, so the interval
in their case at least is a staggering thirty-five hundred years. Moreover, there is much
less material available on comparative Munda, and much of the proto-language has been
reconstructed from material drawn from members of the family outside the subconti¬
nent—there are no etymological dictionaries of proto-Munda, nor of its daughter lan¬
guages. Establishing connections with Munda is made more problematic, since, as Burrow
(1968) noted (to the approval of Masica [1979] and Southworth [1979]), “the evidence
as it is so far established would suggest that these languages in ancient times as well as
now were situated only in eastern India” (328). However, this idea, as will be discussed
later, has recently been challenged by Witzel (forthcoming).

The possibility of Indo-Aryan spontaneously evolving the syntactical innovations and
then influencing Dravidian or Munda, as opposed to being influenced by them, needs
to be excluded as a possibility. Hock (1993) complains that “implicit in the subversionist
view of early Indo-Aryan/Dravidian contact is the assumption of unilateral influence of
Dravidian on Indo-Aryan. . . . No mention is ever made of prehistoric Indo-Aryan in¬
fluence on Dravidian “ (85). Even Emeneau (1980), a committed proponent of Dravidian
bilingualism, is remarkably cautious in his proposals and is forced to acknowledge that
it could have been Indo-Aryan that influenced Dravidian as opposed to being influenced
by it: “Is the whole 1A [Indo-Aryan] history one of self-development, and the complex Dr.
[Dravidian] development something triggered by LA, perhaps even NIA [new IA], influ¬
ence, or, in the case of Kurukh, borrowed from NIA?. . . This . . . solution is less attrac¬
tive than [Dr. influencing IA]. But no easy solution is yet at hand” (174).

As for Brahui, according to Dyen’s lexicostatistical study (1956), “there is a choice
whether Brahui was separated from the other languages by the Indo-Aryan invasion or
whether it represents a migration [from the South], Since a negative migration cannot
be ruled out, the two inferences are equally probable” (625). It is, of course, a perfectly
common linguistic phenomenon for intruding dominant languages to displace preced¬
ing languages, relegating them to small pockets in marginal areas. But it is just as com¬
mon for small groups of language speakers to migrate and survive in small pockets outside
the domain of their own language family; Brahui need not be considered a remnant of
a pre-Aryan, Dravidian substratum, a possibility recognized even by Emeneau (1962b,
70). Like Bloch, Hock (1975, 88) finds the suggestion that Brahui could have emigrated
from the South to the North to be perfectly possible. The Brahui, like the Kurukh and
Malto, actually maintain traditions referring to their external origin. 10 This immigrant
position of Brahui has again recently been asserted by Elfenbein (1987), who finds it
impossible that it could have survived for five thousand years (the Dravidians as will be
discussed below, are supposed to have migrated into India around 3000 b.c.e.) as a
small, isolated linguistic group completely surrounded by alien languages.

All of the linguists who have puzzled over the linguistic innovations in Indo-Aryan
have assumed that the Indo-Aryan speakers were newcomers into India and have, there¬
fore, taken it for granted that “there must have been, from the earliest times, contacts
between the Indo-Aryan invaders and the autochthonous population” (Kuiper [1967]
1974, 141). Other linguistic groups are therefore presumed to have preceded them in
the North of the subcontinent. As Emeneau (1954) puts it, “These invaders did not

84 The Quest for the Origin of Vedic Culture

penetrate into a linguistic vacuum (282).” Approaching the issue on purely linguistic
grounds, those opposing this assumption would have to argue that the areal features of
South Asian languages are independent, spontaneous developments in the languages
concerned and/or that these features have developed as a result of mutual contact other
than that of a substratum nature. This latter possibility would seem to be a perfectly
valid one if only because the South Asian languages are continuing to undergo the process
of convergence. Indeed, some linguists go so far as to suggest that “if the direction of
their development does not change in the future, the now observed tendency to develop
the formal similarity may gain strength and result in the formation of new relationship
ties and of a new language family, which will be neither Indo-European, nor Dravidian”
(Andronov 1968, 13; see also Dasgupta, 1982, 126).

This point, in my opinion, is crucial to this whole issue. If, in die modern and his¬
torical period that can be verified, linguistic convergence is an ongoing process that is
obviously not the result of any bilingual substratum (although adstrata or other types of
bilingualism may certainly be principal factors), then this certainly also could have been
the case in the less-verifiable pre- and protohistoric period. In short, convergence, or
any type of borrowing or similarities between languages, does not necessitate a situation
of a linguistic substratum. 11

Evidence of the Loanwords

Just as the syntactical innovations of Indo-Aryan have been interpreted as evidence of a
pre-Aryan linguistic presence, so have the existence of loanwords, ascribed to either
Dravidian, Munda, and/or unknown origins, been considered proof that the Aryans
imposed themselves on a native populace. Bishop Caldwell presented the first list of
words “probably” borrowed by Sanskrit from Dravidian that he identified according to
certain criteria, most of which hold good today, 12 and which were further refined by
Burrow in 1946. 13

Having applied his criteria, Caldwell found that the Dravidian loanwords he came
up with did not consist of the essential aspects of a vocabulary—the primary words such
as verb roots denoting basic actions, pronouns, body parts, and so on. Such basic terms
are the most durable aspect of a language, even when exposed to major influences from
an alien language family. Caldwell argued that had the pre-Aryan population of North
India been Dravidian, it would have preserved at least some of its own primary Dravidian
terms, which would have resurfaced in at least one or two of the northern vernaculars.
This would especially have been the case in die hypothetical scenario involving a rela¬
tively tiny intrusion of Indo-Aryan speakers superimposed upon a massive population
of Dravidian speakers. If these Dravidian speakers then adopted the language of the
intruders, adjusting it with some of their own structural traits, then surely a few primary
Dravidian words, particularly pronoun forms, also would have survived in some Indo-
Aryan tongue. This was not what Caldwell found: the vocabulary borrowed consisted
of words “remote from ordinary use.”

Soon other lists of Dravidian loanwords in Sanskrit were compiled by scholars such
as Gundert (1869) and Kittel (1894). In 1929, Bloch set the precedent for a practice

Linguistic Substrata in Sanskrit Texts

that continues to the present day: he questioned the Dravidian etymologies for many of
the Sanskrit words that had been proposed by these scholars, fearing that much ety¬
mologizing was not “self-evident” but “a matter of probability and to a certain extent, of
faith” (743). He also made another interesting observation. Many of the Sanskrit words
that were clearly loans had their equivalents in southern Dravidian languages. Whether
or not the Aryans superimposed themselves on a Dravidian substratum in the North is
one issue, but how could the Aryans in the North have been in contact with Dravidians
from the South of the subcontinent? Bloch suggested that either the Dravidian languages
themselves could have been intruders into India, in which case it could have been the
Aryans who borrowed words from the Dravidian speakers who were en route to the sub¬
continent, rather than vice versa, or the words were increasingly borrowed into the written
language from the vernaculars where they were in circulation (743). Since the Dravidian
speakers are also generally considered immigrants into the subcontinent, this sugges¬
tion that loanwords could have been borrowed by Indo-Aryans from invading Dravidians,
as opposed to by Dravidians from invading Indo-Aryans, is another possibility that has
yet to receive scholarly attention 14 (although Witzel [forthcoming] is now also arguing
for a similar possibility).

An indigenous Aryan position might well take interest in Bloch’s suggestion that
the lexical and other features shared by Indo-Aryan and Dravidian (or other languages)
could well have been the result of Dravidian (or Munda) speakers migrating into an
Indo-Aryan speaking area. This possibility will depend on chronology and other issues
and will be discussed further later. Of some relevance, in this regard, is Southworth’s
contention, (1979), based on a comparison of the lists of loans from Dravidian into
Indo-Aryan and vice versa, that “these two lists both seem to suggest a rather wide range
of cultural contacts, and that they do not show the typical (or perhaps stereotypical) one¬
sided borrowing relationship expected in a ‘colonial’ situation, with words for technol¬
ogy and high culture mosdy going in one direction and words for local flora and fauna
mostly in the other (cf. English and Hindi, for example)” (196). In his opinion, “No
picture of technological, cultural, or military dominance by either side emerges from an
examination of these words” (204). This could support the consideration of adstratum
influence between the languages.

Another explanation Bloch considers, which P. Thieme was to find intriguing, was
that individual literary men from the Deccan could have imported Dravidian terms into
classical Sanskrit (in which case, many of the terms would be provincialisms rather than
real borrowings). The massive amount of Sanskrit vocabulary borrowed from the North
by the Dravidian languages, after all, would have been imported by individual Brahmanas
from the North rather than as the result of any major movement of population. This is
a very significant point. These individual Sanskrit-speaking Brahmans who went south
were the cause of extensive lexical adoptions by the Dravidians, but this has nothing to
do with linguistic substrata. The reverse possibility, of individual southerners going north
and importing Dravidian lexica into Sanskrit, seemed to be substantiated for Bloch (1928-
30, 744) by the fact that many of the Dravidian loans did not survive to be inherited by
the later Indo-Aryan vernaculars. 15 The fact that later Pali and Hindi often maintained
the Sanskrit forms rather than the Dravidian ones suggested to Bloch that the Sanskrit
forms had always been die actual words in currency; the non-Aryan foreign synonyms,

86 The Quest for the Origin of Vedic Culture

although appearing in Sanskrit texts, were artificial and temporary innovations. His
remarks seem to be underscored in a more recent study by Masica, who also noticed
that many Dravidian words in Sanskrit had left no living descendants in Hindi. 16

Thieme, in a critique of Burrow, also questioned the assumption that the foreign
words in Sanskrit must have been borrowed from a Dravidian substratum. Burrow, in
a series of articles (1945, 1946, 1983), as well as in an appendix to his book The San¬
skrit Language, had compiled a list of approximately five hundred foreign words in the
Rgveda that he considered to be loans predominantly from Dravidian. Thieme (1955)
like Bloch, is much more inclined to ascribe most of them to borrowings from the ver¬
naculars that increasingly crept into the more elite Brahmanical circles eventually be¬
coming Sanskritized. He remarks:

The bearers of the sacred language are obviously and professedly eager to keep their speech
pure and unadulterated. . . . We have no evidence for speech contact with Dravidian-
speaking people in the North of India in olden times: no Patanjali, Yaska gives us . . .
any such indication. ... all the Dravidian languages known to us fairly bristle with loans
from Sanskrit and the Aryan vernaculars. Dravidian literature in South India came into
existence under the impulse and influence of Sanskrit literature and speech. Wherever
there is a correspondence in the vocabularies of Sanskrit and Dravidian, there is a pre¬
sumption, to be removed only by specific argument, that Sanskrit has been the lender,
Dravidian the borrower. ... 1 should consider it likely that. . . loans had first been given
a homestead in a vernacular and penetrated thence into Sanskrit. If Patanjali (Mahabh.

1.2.7-9) looks on a Magadhi word as a ‘barbarian’ ( mleccha ) how much more would he
have despised a word of a completely different language? (438-439)

That no native grammarians have recorded any awareness of a non-Sanskritic indig¬
enous population (although drey do draw attention to des'ya words) is significant. Thieme’s
point is further bolstered by the fact that, to my knowledge, none of the earliest foreign¬
ers in India who left written records—the Greeks or the Chinese—made any reference to
a subjugated population speaking a different language from the ruling elite, whether in
domestic situations or in marginal geographic areas. Thieme is proposing that whatever
words were undeniably loans could have infiltrated into the northern vernaculars some¬
how or other, perhaps as a result of individuals bringing them from the South, from
where they surfaced into Sanskrit, just as loans circulate in any language. He then pro¬
ceeds to give alternative etymological explanations for some of the words Burrow had
listed, preferring to derive them from vernaculars or from Indo-European word forms
rather than from Dravidian. 17 In 1994, he again challenged many supposed Dravidian
etymologies, this time from Mayrhofer’s Etymologisches Worterbuch, stating that “it is
. . . quite legitimate to consider the possibility of Sanskrit borrowing from any non-Aryan
Indian language. Yet, if a word can be explained easily from material extant in Sanskrit
itself, there is little chance for such a hypothesis” (327). His doubts about many of the
alleged Dravidian etymologies have been supported by scholars such as H. H. Hock
(1975, 1984d) and R. P. Das (1995).

The controversies between the “Dravidianists” and “Indo-Europeanists” on the ori¬
gin of unusual Vedic and Sanskrit etymologies, with a subsequent rise of an “Austro-
Asiatic” contingent, has been a strongly contested issue. Thieme characterizes the activi¬
ties of some of his colleagues as a “zeal for hunting up Dravidian loans in Sanskrit”
(1994, 327). From the other side, the advocates of a major Dravidian (and Munda)

Linguistic Substrata in Sanskrit Texts 87

component in Sanskrit lexemes are not impressed by the “resort to tortuous reconstruc-
tions in order to find, by hook or by crook, Indo-European explanations for Sanskrit
words” (Burrow 1956, 321 ). 18 Kuiper (1991,91) feels such “dogged resistance” needs
to be understood through the “psychological background,” which he traces back to the
days when Sanskrit was the undisputed model language for comparative Indo-European
studies. As far as he is concerned, the very fact that satisfactory etymologies have not
been found after a century and a half of etymologizing is sufficient evidence in itself to
suspect the Indo-European pedigree of a word.

The fiercest controversy has naturally revolved around the oldest text, namely, the
Rgveda since this preserves the earliest “real” linguistic evidence. Kuiper (1991) has laid
claim to 380 foreign words, or 4 percent of die Rgvedic vocabulary (95). 19 This is dra¬
matically ambitious in comparison to any other previous claim. Of special interest in
his lists are at least thirty-five non-Aryan names for individuals, families, and tribes—
including several, such as the dasa Balbutha Taruksa, who were patrons of priests and
therefore participated in Vedic culture (6-8). 20 Witzel (1999a) reckons that almost half
of the fifty-odd tribal and clan names in the Rgveda have no Indo-Aryan or Indo-Euro¬
pean etymology. 21 Kuiper (1991) finds “clear traces of an influx of non-Aryan beliefs”
that are “not so much a case of borrowing ... as rather an echo of a foreign religion
being incidentally audible in die circles of the Rigvedic rsis” (16). In contrast to Southworth
and Hock, he interprets the linguistic evidence as indicating that the social interaction
took place at the lower end of the social echelon, not the highest. 22 He holds that these
words have been borrowed from Old Dravidian, Old Munda, and several other lan¬
guages. It is also important to note that most of the foreign words in the Rgveda and
Atharvaveda are rare or hapax legomena. In addition to lexical and cultural influence, he
elaborately documents traces of patterns of foreign influences in phonology, morphol¬
ogy, and syntax, as well as the adaptation of foreign phonemes. Kuiper’s innovative and
meticulous work is the result of perhaps half a century of research in this area.

Emeneau (1980) in contrast, would seem to be relieved if scholars would agree on
even one single loan: “If any such words can be found, even one or two, they will pro¬
vide secondary evidence [of a Dravidian substratum). ... I can only hope that the evi¬
dence for mayuura —as a RV borrowing from Dr. is convincing to scholars in general"
(179-183). Hock (1996b) is reluctant to allow even this, noting that the Dravidian cog¬
nate mayil occurs in southern Dravidian languages, far from the northwestern lands of
the Rgveda (38). Since Dravidian has a similar root denoting both ‘cat’ and ‘peacock’,
and Sanskrit marjaraka means both ‘cat’ and ‘peacock’, Hock prefers to consider the
word onomatopoeic. 23 In any event, linguists are quick to warn that “evidence for sub¬
stratum cannot rest on an isolated lexical item, but must be based on a coherent pat¬
tern” (Salmons 1992, 267). In this regard, it might be worth noting Hamp’s comments
(1990b) regarding substratum identification: “We cannot fasten upon single facts. This
means that we cannot seize upon a single isolated personal name. . . . we must have
homomorphous distributions of multiple sets of equivalencies. It is the same formal
requirement, of course, that we insist on for demonstration of genetic Lautgesetze.” The
foreign features attributed to a substratum must be consistently perceptible over a range
of data: “For acceptable substrata, when we finish stating these equivalencies we must
make a cohesive statement of structural features which are to be assigned to those sum¬
maries of homomorphous equivalent sets. Only then can we say we have a substratum

The Quest for the Origin of Vedic Culture

worth talking about. Until those requirements have been fulfilled you may have an
interesting breakfast observation, but you do not have an argument” (293). Establish¬
ing such consistent features is exactly what Kuiper has attempted to do.

Kuiper’s work, in addition to compiling a list of loanwords, has attempted to iden¬
tify systematic patterns of non-lndo-Aryan linguistic features in the Rgveda, particularly
in phonology and morphology, that can contribute to identifying features typical of the
underlying foreign language(s) that influenced Sanskrit. This aspect of his work is more
problematic to the Indigenous Aryan position than his list of foreign lexica. Many of
these features, such as some of the prefixes, point to Munda. These “are unknown in
Dravidian but were common in Austro-Asiatic. . . . According to some scholars Munda
was never spoken west of Orissa, Bihar, Madhya Pradesh and eastern Maharashtra. . . .
The obvious occurrence of Old Munda names in the Rigveda points to the conclusion
that either this statement should be revised or that some parts of the Rigveda . . . stem
from eastern parts of North India” (Kuiper 1991, 39-40). Michael Witzel has predi¬
cated more far-reaching conclusions on Kuiper’s observations, which will be discussed
later.

Nonetheless, Kuiper’s etymologizing has been sharply criticized by Rahul Peter Das,
who, in keeping with his predecessors, challenged the “foreignness” of many of Kuiper’s
non-lndo-Aryan words. Das (1995) warns that one should not assume that “problem¬
atic words are foreign—they might be, but they need not be, for not being able to find
a clear Indo-European etymology does not automatically imply that an Indo-European
origin is impossible." As far as he is concerned, there is “not a single case in which a
communis opinio has been found confirming the foreign origin of a Rgvedic (and prob¬
ably Vedic in general) word.” The considerable differences of opinion “may be due to
the fact that many of the arguments for (or against) such foreign origin are often not the
results of impartial and thorough research, but rather of (often wistful) statements of
faith” (208; italics in original). Das is here echoing the acknowledgment of Emeneau
(1980) that vocabulary loans from Dravidian into Indo-Aryan are “in fact all merely
‘suggestions.’ Unfortunately, all areal etymologies are in the last analysis unprovable,
are ‘acts of faith’. ... It is always possible, e.g., to counter a suggestion of borrowing
from one of the indigenous language families by suggesting that there has been borrow¬
ing in the other direction” (177). Das (1995), who accepts the external origin of Indo-
Aryan on grounds other than the substratum hypothesis, points out that there is “not
a single bit of uncontroversial evidence on the actual spread of Dravidian and Austro-
Asiatic speakers in pre-historic times, so that any statement on Dravidian and Austro-
Asiatic in Rgvedic times is nothing but speculation” (218). He reiterates the caution
that the material being compared is separated by vast amounts of time and distance,
often with scant regard being paid to the need for extensive philological investigation in
establishing the exact semantics of dubious words, particularly in the case of Dravidian
and Austro-Asiatic.

One study conducted in 1971 gives an indication of the extent to which scholars can
disagree. In his doctoral thesis at Poona University, A. S. Acharya took all the Sanskrit
words that had been assigned a Dravidian etymology over the years, resulting in a list
of over twelve hundred words said to be borrowings from Dravidian. 24 He then searched
for these words in Mayrhofer’s etymological dictionary and found that Mayrhofer had
accepted only 25 percent of the words as being Dravidian with any degree of certainty

Linguistic Substrata in Sanskrit Texts 89

(Acharya 1971, 15-74). Along the same lines, he extracted all the Sanskrit words that
had been assigned a Munda derivation by scholars and arrived at over one thousand
different words. 25 Here, too, he found that Mayrhofer had accepted only 21 percent of
these words as borrowings from Munda that had any degree of certainty. When we
bear in mind that scholars such as Thieme and Hock consider even Mayrhofer’s Dravidian
etymologies to have been awarded far too liberally, the differences of opinion are con¬
siderable. Such problems are not unique to South Asia: Salmons (1992, 266) refers to
scholars who have attempted to deny the role of substratum on Germanic on the grounds
that patient etymological work could eventually reveal Indo-European forms for most of
the items in question.

On a different note, there are also methodological problems with the often quoted
idea—outlined, for example, by Burrow (1968c, 326)—that there seems to have been a
small number of foreign words in the Rgveda, which increased marginally in the other
Vedas, grew considerably in the later Vedic literature, and peaked in the Epics and
Puranas, before dwindling in the Prakrits and new Indo-Aryan languages. These data fit
neatly with the generally accepted scenario of pre-Vedic-speaking Indo-Aryans intruding
into an area inhabited by speakers of other languages. It suggests that by the time the
Aryans had composed the Vedic hymns only a few indigenous words had permeated
the texts. As the process of bilingualism developed (involving both the indigenes in the
north of the subcontinent preserving some of their native lexicon as they adopted the
Aryan languages, and post-first generation Aryans themselves utilizing non-Indo-Euro¬
pean words as they merged with the local people), the loanwords increased, namely,
during the Epic period. Finally, there were no more bilingual speakers left—everyone
had adopted a form of later Indo-Aryan—at which time the appearance of foreign lexemes
appears to cease as evidenced by the decrease of loans in the later Indo-Aryan texts.

This idea of the number of loanwords in Sanskrit increasing and then decreasing
seems to be methodologically untenable. First of all, whose lists do we go by—Kuiper’s
(380 loans in the Rgveda) or Thieme’s (no loans at all)? Second, even if we allow that
the number of loanwords in Sanskrit did increase in progressively later texts, it is mis¬
leading to conclude that this was the result of more loans filtering through. The num¬
ber of loans has to be calculated as a ratio of the total number of words in the text, and
the percentage of foreign words in earlier texts has to be compared with the percentage of
foreign words in later texts to determine whether the number of loans was increasing,
decreasing, or, as is quite likely, more or less constant. To my knowledge, no such study
has been attempted. Moreover, Emeneau (1980, 184) recognized that the claim that
New Indo-Aryan had ceased to borrow was erroneous, a conclusion he formed simply
by looking at the number of New Indo-Aryan entries in Turner’s Comparative Dictio¬
nary of Indo-Aryan Languages that had been assigned a Dravidian origin. Masica (1979)
found that the “non-Dravidian, non-Munda element in the Indo-Aryan lexicon persists,
and even grows” (138). There are large numbers of such foreign words in the modern
Indie languages. As I stressed previously, the syntactical and lexical convergence of the
South Asian languages is continuing and increasing to this very day to such a degree
that some scholars have suggested that a new language family is developing in South
Asia, distinct from both Indo-European and Dravidian. This continual process of ac¬
cepting foreign words and syntactical features (possibly even at an increasing rate) in
New Indo-Aryan is not the result of any foreign linguistic substratum, so it is perfectly

90 The Quest for the Origin of Vedic Culture

legitimate to ask why this had to be the case in the protohistoric past. If adstratum can
account for such interactions in the present, one would need to produce a convincing
explanation for why this could not have been the case in the protohistorical period.
Moreover, even if for argument’s sake it could be established that the percentage of
foreign words in the Rgveda was considerably less than that of later texts, the explana¬
tion could well be a sociolinguistic (i.e., cultural) one. Many scholars such as Thieme
have drawn attention to the linguistic puritanism of the Vedic texts. These are sacerdo¬
tal hymns describing ritualistic techniques that were preserved by a culturally distinct
group of specialists who, like any elite, took pains to isolate their speech from common
vulgarisms. The Epics and Puranas deal with the real world—tribes, geography, history,
intrigue, war, and the religion of the people. Naturally, the vocabulary in these latter
texts will be far less conservative and more representative of the language of the street,
so to speak. One is not compelled to interpret any possible disparate proportion of
non-Aryan words in different genres of texts as proof of a linguistic substratum.

Terms for Flora in Indie Languages

Nonetheless, even when the most generous allowance is made in favor of the Indo-
Europeanists, few would deny that there are many words borrowed from non-Aryan
languages in Sanskrit texts. Dravidian and Munda are not the only languages that have
been considered as possible sources: the difficulty in tracing the etymologies of many of
these words has caused most linguists who specialize in this area to propose that many
of the features and loans in Sanskrit must have come from languages that have disap¬
peared without a trace. 26 Burrow (1956) goes so far as to suggest that “it may very well
turn out that the number of such words which cannot be explained will outnumber
those which can be. This is the impression one gets, for instance, from the field of
plant names, since so far only a minority of this section of the non-Aryan words has
been explained from these two linguistic families” (327).

This point is significant, particularly if these are local plants native to the Northwest.
If the Aryans were indigenous to India, why would they have borrowed names of native
plants from other language families instead of possessing their own terms composed from
Indo-Aryan roots or derivatives? 27 This discrepancy would be natural if the Aryans were
newcomers to the subcontinent, in which case they would readily adopt the names of
unfamiliar local fauna and animals current among the native population. Kuiper (1991)
finds that the words from his list—which contain about a dozen words that can be directly
connected with agriculture—“testify to a strong foreign impact in almost every aspect of life
of an agrarian population. ... an (originally) non-Aryan agrarian population was more or
less integrated into a society of a predominandy different character” (15).

The “different character” referred to by Kuiper reflects the common opinion among
scholars that the economy of the Indo-Aryans is one of predominantly nomadic pasto-
ralism. This, to a great extent, is predicated on the fact that the Indo-Iranian terms for
agriculture were different from the set of corresponding terms shared in the western
Indo-European languages. Scholars since at least the time of Schrader (1890) have con¬
cluded from this that “it becomes impossible to doubt that the Indo-Europeans, when
they made their first appearance in history, were still possessed with nomadic tenden-

Linguistic Substrata in Sanskrit Texts 91

cies” (282). From such a nomadic genesis, the western branches and eastern branches
subsequently encountered agriculture after they had gone their separate ways to the west
and east, respectively, and borrowed or coined their respective sets of different terms
independently. The Indo-Aryans, from this perspective, were still nomadic when they
arrived in South Asia, although they practiced nonsedentary agriculture (i.e., they en¬
gaged in agriculture for periods of the year, according to seasonal and other factors,
before moving on). Most Vedic philologists believe that this hypothesis is supported by
the economic culture reflected in the Vedic texts. Nowadays, this lexical difference be¬
tween the eastern and western indo-European languages is explained by the polycentric
origin of agriculture from two or three food-producing centers (Makkay 1988; Masica
1979). Moreover, Indo-Europeanists (Diebold, 1992; Mallory, 1997) do argue that the
proto-Indo-Europeanists were agriculturists.

Southworth (1988) determines that the agriculture of the Vedic Aryans was appar¬
ently limited to barley and beans in the earliest texts. Rice and sesame appear in later
texts, and wheat later still: “These facts support the view that the earliest Vedic texts
were associated with a mountain-dwelling, primarily herding people who were unac¬
quainted with the type of floodplain agriculture practiced by the Harappans.” The fact
that some of the products cultivated in the Indus Valley—wheat, cotton, sesame, dates,
and rice—are absent from the earliest Vedic texts “is evidence for a lack of substantial
contact between these people and the Harappans” (663). The assumption here, which
must be kept in mind, is that the fact that the Rgveda happens not to mention rice,
wheat, and so on, indicates that these items were unknown to the earliest Indo-Aryans.
Southworth’s conclusions should be considered, provided one keeps in mind that the
text is not a compendium of agricultural terms; one must be wary of drawing too far-
reaching conclusions based on argument! ex silentio , 28

As 1 will discuss more fully later, the borrowing and coining words for flora and
fauna peculiar to India, as opposed to possessing ‘primitive’ terms for them (i.e., their
possessing an Indo-European root) generally indicates to linguists that the item in ques¬
tion is new and unfamiliar to the speakers of the Indo-Aryan languages. In his research,
Masica found 80 percent of the agricultural terms in Hindi to be non-Aryan—55 per¬
cent of which were of unknown origin (some inherited from Sanskrit). What is espe¬
cially important to note here is that out of the total number of items in the survey, only

4.5 percent were Austro-Asiatic—some of which Masica acknowledges could have been
borrowings from Indian contacts with the Mon-Kmer peoples in Southeast Asia rather
than from the Kolarians of India proper (Masica 1991, 129-139).

This insignificant number caused him to concur with Burrow’s opinion that the
Munda languages could not have been present in the Northwest of India in prehistoric
times. Had they been a principal component in any pre-Aryan linguistic substratum,
the number should have been far greater. The same would apply to Dravidian’s poor
showing. Masica’s study (1991) presents a further obstacle to theories proposing a
Dravidian linguistic substratum in North India—only 7.6 percent of agricultural terms
in Hindi have Dravidian etymologies. Moreover, “a significant portion of the sug¬
gested Dravidian and Austroasiatic etymologies is uncertain” (1 34). Thus, 4.5 and

7.6 percent are generous. This lacuna forces him to wonder whether the Indus Valley
might not have been multilingual and, in concordance with numerous linguists since
the time of Caldwell, to postulate the existence of another unknown language (or

92 The Quest for the Origin of Vedic Culture

languages) existing as linguistic substrata in Indo-Aryan times (138). He labels this
tongue “language x.”

Southworth (1979) is also specific in noting that the terms for flora did not come
from Dravidian. Scouring the work of Burrow and Emeneau, he extracted fifty-four words
for botanical terms, including trees, cereals, edible gourds, spices, and beans, that have
been considered loans common to both Indo-Aryan and Dravidian. So Dravidian was
not the lender for these botanical terms. Nor was Munda: out of these loans, Southworth
(1979) finds only five that are shared with Munda, causing him to suggest that “the
presence of other ethnic groups, speaking other languages, must be assumed for the
period in question” (205), with hardly “the slightest hints” as to what these languages
might have been.

Curiously, in contrast to Masica’s results showing that 80 percent of the agricultural
vocabulary in Hindi was non-Indo-Aryan, a study conducted by Wojtilla (1986) using
the material offered by a few special vocabularies of agricultural terms in the Hindi belt,
coupled with his own collection of terms from the vicinity of Varanasi, seemed to pro¬
duce conflicting results. Wojtilla found that “the agricultural vocabulary so collected
mostly consists of tatsama and t adbhava words already known in Sanskrit and Prakrits”
(28). The article does not discuss why this finding contradicts Masica’s (which Wojtilla
uses as a model), but this suggests that further extensive analysis of this issue is in
order. Wojtilla does concur with Masica on another point, however, which is that
“Dravidian influence is less than has been expected by specialists” (34).

The matter of loanwords is another area almost completely overlooked by Indigenous
Aryanists. S. G. Talageri is the only scholar I have encountered who attempts to ad¬
dress some of the issues connected with linguistic substrata, but he seems unaware of
most of the work that has been done in this area. His sole source seems to be Chatterji’s
Origin and Development of the Bengali Language, wherein there is a short list of about
forty words of “probable Dravidian origin” (Talageri 1993, 42). Talageri searches for
each item in Buck’s dictionary of synonyms and finds about 20 percent of them have
been assigned Indo-European derivatives. 29 He then proceeds to try his hand at estab¬
lishing a few Sanskrit etymologies for some of the remaining words, 30 finally conclud¬
ing that “the overwhelming majority of Sanskrit names for Indian plants and animals
are derived from Sanskrit and Indo-European roots” (205). This is a rather sweeping
conclusion that does not seem to be based on an awareness of most of the basic mate¬
rial in this area—over a century of research in the area of Sanskrit loanwords deserves a
less cursory dismissal.

Talageri (1993) does accept the existence of some loanwords from Austro-Asiatic and
Dravidian and accounts for their existence by contending that “the Dravidian languages
were always spoken in South India, . . . the Indo-European languages were always and
originally spoken in North India, and the Austro-Asiatic languages were always spoken
in north-eastern and east-central India” (206). He remarks that many of the plants and
animals must have had geographically specific areas of cultivation and natural habitats
so that “if the common name for any Indian plant is proved to be of Austric or Dravidian
origin, it will help in locating the part of India in which the plant had its origin” (206).
Although no language can have existed anywhere in the world since all eternity, the
point is an interesting one: Masica (1979) remarks that “there is really no Indian agri¬
culture as such, but a group of related regional complexes differing in important details,

Linguistic Substrata in Sanskrit Texts 93

including inventories of cultivated plants. Sanskrit, being a supraregional language,
incorporates terms reLating to various regional features” (58). Since Talageri seems largely
unaware of the significant number, or etymological nature, of the non-Aryan words for
plants and animals found in Sanskrit texts, he offers only a single example in support
of his premise (eld ‘cardamom’ front the Tamil, a spice thought to have originated in
Kerala).

Nonetheless, this method, pursued with methodological rigor as Southworth (forth¬
coming) has done, will be an important contribution in this area. Agricultural terms are
an essential part of the data concerning Aryan origins that Indigenous Atyanists have
to address. As it stands, many of the words in Sanskrit for domestic animals and their
products are generally accepted as Indo-European derivatives, but few agricultural or
botanical terms are. Masica’s study found a significant percentage of Hindi words con¬
nected with animal husbandry were Indo-European or derivatives thereof, but fewer
cereals, pulses, roots, fruits, and vegetables could be accounted for in this way. Most
scholars have quite reasonably inferred from such data that pastoral nomads entered
into the subcontinent with their culture of livestock herding and encountered strange
local flora whose names they had to borrow from the indigenous people. Such evidence
is an important ingredient in the Aryan Invasion hypothesis. To my mind, the non-
Indo-Aryan nature of the words for the flora of North India is one of the few truly
compelling aspects of the entire substratum theory.

However, even in this regard, the existence of a pre-Indo-Aryan linguistic substra¬
tum does not have to be the only explanation for the many botanical terms in Sanskrit
that do not have Indo-European etymologies. Scanning the gamut of Sanskrit texts from
different chronological periods, Southworth (forthcoming) finds that from a total 121
terms for plants, only a little over a third have Indo-European etymologies, and an ad¬
ditional third have unknown etymologies. First of all, there is much more work to be
done in scrutinizing the etymologies and compositions of problematic words—we have
seen how dramatically scholars have disagreed with each other. The task is a daunting
one, however, and a large proportion of words are likely to remain untraceable. With
regard to such recalcitrant terms, because foreign botanical items (millet, sorghum, etc.)
have been continually imported into the subcontinent since time immemorial, it is more
than probable that some have maintained their original foreign names. In many cases,
these non-Indo-Aryan designations could be traceable to other language families, and
the linguistic history of such words could tell us much about the origin of their refer¬
ents. In this category of words, then, it is the plant, not the Aryans, that would be the
intruders to the subcontinent. 31 In addition, the same basic possibilities outlined ear¬
lier for Dravidian and Munda linguistic relationships need to be considered: to what
extent can these unknown items ascribed to “language x” be the result of loans, or
adstratum relationship between Indo-Aryan tribes and other unknown ones, rather than
the result of substratum. 7

Even within the subcontinent itself, plants and vegetables need to be correlated with
their areas of origin to see whether their names can be connected with the linguistic
groups known to have resided there. Paleobotany has a potentially significant role to
play in this type of “linguistic archaeology.” The problems involved, however, are daunt¬
ing—as Polome (1990b) notes, “one is rather reluctant to extrapolate from relatively recent
data to archaic ecology” (276). Southworth has taken some significant steps in this re-

94 The Quest for the Origin of Vedic Culture

gard. His forthcoming book Linguistic Archaeology of the South Asian Subcontinent
combines paleobotany with etymological analysis in an attempt to illuminate the his¬
torical relationship between plants and human societies in South Asia. He provides
extensive lists of plants known in Sanskrit texts, determines their etymologies where
possible, and identifies their probable places of origin.

Southworth lists six plants that are believed to have come from different parts of the
African subcontinent (finger millet, sesame, bulrush millet, sorghum, cowpea, and okra).
He finds that another six names with Dravidian etymologies refer to plants whose ori¬
gins lie to the east of India, suggesting that they may have been transported by sea to
peninsular India by Dravidian speakers (Mahdi’s study [1998] on the transmission of
Southeast Asian cultigens also finds that, with some exceptions, “the plants and crops
from Southeast Asia acquired new names in the process of transmission to India” [411]).
Southworth further identifies seven items (karpasa ‘cotton’; kahguni ‘foxtail millet’; kadala
‘banana’; tambula ‘betel’; nimbu ‘lemon’; marica, ‘pepper’; and sarkard ‘sugarcane’) that
are believed to have originated in Austroasiatic languages. Most of these too, predict¬
ably, have their domestic origin in the east of India. In short, we have a long history of
plants being imported into the subcontinent.

More important, Southworth introduces the category of plants that may have origi¬
nated in India. From these he identifies nine plant terms shared by Old Indo-Aryan
and Dravidian where the direction of borrowing is not clear. These suggest to him that
the terms may have been borrowings into Indo-Aryan and Dravidian, both of which he
considers to be intrusive, from one or more previous, indigenous languages. However,
the actual point of origin of most of these plants either is uncertain or occurred in the
South (cardamom) or the Northeast (mango), in areas where die historical existence of
non-Indo-Aryan languages is not under dispute. It is plants that are native to the North¬
west that are critical for this discussion.

Apart from these, Southworth’s work clearly reveals a history of plant importation
into the Indian subcontinent. However, the importation of foreign plants need not denote
the foreignness of Indo-Aryan speakers. Indo-Aryan speakers in India still to this day
import and cultivate new crops and retain their foreign names, as they have done through¬
out history. One need only go to one’s local supermarket to experience this principle:
exotic fruits from exotic countries are imported into our societies (and sometimes even
transplanted and grown locally) while nonetheless retaining their original foreign names,
which soon become part of our own vocabularies. Therefore, although the foreign names
for flora may very well be indicative of a pre-Indo-Aryan substratum, this need not be
the only explanation; these terms could simply be loans denoting items imported into
a preexisting Indo-Aryan-speaking area. Only the etymologies of terms for plants indig¬
enous to the Northwest of the subcontinent have the potential to be conclusive. If the
Indo-Aryans were native to the Northwest, one would expect Indo-Aryan terms for plants
native to the Northwest. If such plants could be demonstrated as having non-Indo-Aryan
etymologies, then the case for substratum becomes compelling. However, Southworth’s
lists show no instance of plants native to the Northwest that have non-Indo-Aryan ety¬
mologies.

This is the essential point. If the Aryans were indigenous to at least the Northwest
of the subcontinent, one would expect that there should be a higher percentage of, if
not Indo-European (since, as will be discussed in the next chapter, items unique to India

Linguistic Substrata in Sanskrit Texts

would not be expected to have cognates elsewhere), at least Indo-Aryan-derived names
for plants known to be common in the area inhabited by the compilers of the Rgveda
(provided these terms exhibit permitted Indo-European forms). As a side note, it is also
important to repeat that if the etymological obscurity of plant names in Sanskrit texts is
to be considered detrimental to the case of the Indigenous Aryanists, it is equally detri¬
mental to the case of anyone promoting Dravidian (or Munda) as the indigenous pre-
Aryan language of the Northwest. Most of the plant names are not traceable to Munda
or Dravidian either (although, of course, Dravidian or Munda could have preexisted
Indo-Aryan and passed into Sanskrit terms that they had borrowed from the “language
x” that preceded them, in turn).

Also of particular relevance are the etymologies in Sanskrit texts for the terms for
plants that have been found in the archaeological record of the Northwest in strata that
date prior to when the Indo-Aryans are supposed to have entered the subcontinent,
namely, before the second millennium b.c.e. Southworth’s study lists six plants from
the pre-Harappan and Early Harappan period: yava ‘barley’; tula ‘cotton’; kharjura ‘date’;
draksa ‘grape’; badara ‘jujube’; and godhitma ‘wheat’. All of these plant products were
found in Mehrgarh, Baluchistan, around the sixth millennium b.c.e. except grape (Kash¬
mir, late third millennium b.c.e.) and jujube (Mundigak in Baluchistan, fourth millen¬
nium b.c.e.). From these, only barley has a clear Indo-European etymology. Godhitma
(lit. ‘cowsmoke’, godum in Dravidian) seems to be a folk etymology, which Witzel con¬
siders a Sanskritization (and Dravidianization) of a Near Eastern loanword (Proto-Semitic
* hant, Old Egyptian xnd). Four of the terms seem etymologically unaccountable from
Indo-European or Indo-Aryan roots.

This evidence is problematic for the Indigenous point of view. If the Indo-Europe¬
ans had come from the Northwest of the subcontinent, one would expect that the plants
cultivated there since the sixth millennium b.c.e. would have Indo-European etymolo¬
gies, which would then have evolved into Indo-Aryan forms. The only way to otherwise
account for the four items with foreign etymologies might be to argue that since all of
these plants (with the exception, perhaps, of dates) might have been imports into South
Asia (i.e., they have been found in earlier archaeological contexts outside the subconti¬
nent), the original foreign names from their places of origin could have been retained
throughout prehistory. After having been transmitted down through die centuries, such
names eventually surfaced in the Vedic texts as foreign words, or were assigned folk
etymologies.

Words consist of roots, suffixes, and endings (the word singers has a root sing, a
suffix -er, and an ending -s). A new formation or coinage refers to a word that is San¬
skrit in form (i.e., with known Indo-Aryan morphological units such as suffixs or pre¬
fixes) but that either does not contain as Indo-European root or contains morphological
units that are Post Indo-European. Such later developments do not necessarily reveal an
ancient Indo-European etymology but might suggest recent coinage and therefore immi¬
gration into a new area (although new formations do not always denote substratum,
since there is nothing preventing indigenous people from continually coining new words
that reflect the linguistic developments extant at different chronological periods). The
safest way of determining whether a word is a loan is from the root. Briefly, Indo-Euro¬
pean roots are of the forms (s)(C)CeC(C/s), where C is a consonant, ( ) is an optional
consonant -C- is a standard Indo-European vowel (in the ablaut series e/o/0/e/o out-

96 The Quest for the Origin of Vedic Culture

lined in chapter 3), and s is a sibilant. There are, however, certain limitations in the
consonant combinations of the roots. The following combinations, for example, are
generally not tolerated in an Indo-European root: two voiced consonants (deg), unvoiced
and aspirated consonants (tegh or dhek) and two identical consonants (pep), Also, the b
consonant is very rare. A word not fitting the basic Indo-European root pattern is an
immediate and obvious candidate for being a non-Indo-European loan. Thus, the name
Balbutha is a clear non-Indo-European, non-Indo-Aryan term. A few points should be
borne in mind before insisting that unfamiliar botanical words in Sanskrit are, of ne¬
cessity, proof of a non-Aryan linguistic substratum. As Masica (1979) points out from
his study of agricultural terms in Hindi: “It is not a requirement that the word be con¬
nected with a root, of course: there are many native words in Sanskrit as in all lan¬
guages that cannot be analyzed’’ (61). This inscrutability of certain terms is especially
prominent in terms for flora. C. D. Buck (1949), who compiled a dictionary of syn¬
onyms in the principal Indo-European languages, remarks that for most Indo-European
trees, “the root connections are mostly obscure” (528). Likewise, the same applies to die
inherited names of animals (135). Friedrich (1970), in his study on Indo-European
trees, found only three roots that could be “cogently connected with a verbal root. . . .
the great majority of PIE tree names were . . . unanalyzable nominal roots, and . . .
for their reconstruction the most relevant branches of linguistics are phonology and
semantics” (155).

The general inscrutability of terms in Indo-European is an important point: Sanskrit
words for plants and animals do not automatically have to be considered foreign and
rejected as possibly being Indo-Aryan due to dubious derivation because obscure ety¬
mological pedigrees would appear to be the norm for most plant and animal terms in
Proto-Indo-European in general (this could, of course, be explained by postulating that
the Proto-Indo-Europeans were themselves intrusive into whatever area was their home¬
land prior to their dispersal and borrowed terms for fauna and flora from the preexist¬
ing substratum in that area). 33 Talageri (1993) comments in this regard that “unless
one is to presume that the Proto-Indo-Europeans were not acquainted with any animals
or plants at all, one has to accept that etymologically obscure names may be ‘what were
at first colloquial or even slang words,’ and that etymological obscurity need not neces¬
sarily indicate a non-Indo-European source unless such a source . . . can be specifically
. . . demonstrated” (206). One must be cautious of too-hastily branding a word as
non-Indo-European simply because one has not been successful in establishing an
Indo-European etymology. Talageri also makes the caveat, in the undeniable instances
in which a term for an item is demonstrably a loanword, that such borrowing does not
necessarily indicate a lack of prior acquaintance with that item; there may be cultural or
other reasons for the adoption. Masica (1979) illustrates this point by remarking that
the foreign name for an item may replace an older indigenous name—French ‘pigeon’
has replaced the older Germanic ‘dove’ (61) in English, for example, but this has noth¬
ing to do with linguistic substrata or with prior ignorance of the object. Folk etymolo¬
gies can also replace older terms—Vedic ibha, for example, has been replaced by the
popular folk etymology of hastin— ‘the one with a hand’. On the other hand, foreign
words can be made to appear indigenous by Sanskritizing them or assigning a new
Sanskritic name to their referent—a practice Sanskrit grammarians were expert at. In

Linguistic Substrata in Sanskrit Texts 97

such an instance, a word derived from Indo-Aryan may be a later gloss over an original
non-Aryan term.

As an aside, one wonders whether, if all branches of the Indo-European language
family (Balto-Slavic, Italic, Germanic, etc.) had preserved ancient corpora of texts dating
back to at least the early second millennium b.c.e., one might riot have found that they
all preserved evidences of foreign floral, faunal and other typical indicators of substrata.
Where would one have placed the homeland if the areas where all the Indo-European
languages are spoken were to be eliminated by the logic used to eliminate the North¬
west of the subcontinent? While this may well be an unwarranted flight of fancy, it
seems fair to point out that the homeland candidacy of the Volga Valley steppes, for
one, is actually advantaged by the absence of ancient textual sources in the Indo-Euro¬
pean languages spoken in that area (such as Balto-Slavic) that might well have proved
detrimental to their case were they to have been preserved and discovered. The same
holds true for other postulated homelands.

Returning to the Rgveda, Bloch’s and Thieme’s proposals also deserve to be kept in
mind—many peculiar words are quite likely to have had their origin in the Prakrits or
other “low” culture vernaculars. This is especially pertinent in the case of plants and
other agricultural terms, since such words would have been the daily subject matter of
the tribes and “lower” social groups who tilled the soil, gathered the flora and herbs,
and dealt with animals. These tribes may have picked up foreign plants and their terms
from their wanderings and trade interactions with other language groups. Kuiper (1955),
who has classified the foreign words according to the various spheres of human life in
order to estimate their general character, found that “the vast majority of the Rigvedic
loan words belong to the spheres of domestic and agricultural life. They belong not
only to the popular speech . . . but to the specific language of an agrarian population”
(185). Although Kuiper sees this population as one preceding tire Indo-Aryan-speaking
one, the issue at stake is how to preclude the possibility that these people might always
have been speakers of Indo-Aryan dialects, albeit saturated with a ‘deshi’ folk lexicon,
much of it etymologically unexplainable with our present resources. Sociolinguistics is
likely to have a role to play here. These may not be the types of people likely to be over¬
concerned about preserving pristine speech forms (which could well explain why plant
and animal forms in general are etymologically indeterminate). Nomadic tribes, per¬
haps trading animal and faunal products between different regions and language groups,
easily could have been the bearers of loanwords connected to the merchandise they
bartered. As discussed earlier, flora and fauna are precisely the types of items that are
continually imported into new environments to this very day, often retaining their for¬
eign names. Of course, certain things are more likely to be imported than others: edible
items or flowering blossoms would likely be more amenable to trade than trees, for
example, but this possibility has not recieved adequate attention.

There is ample evidence of foreign personages and tribes in the Vedic period. Kuiper
lists some twenty-six names of Vedic individuals who have non-Indo-Aryan names, with
which Mayrhofer concurs (Kuiper 1991, 6-7). Witzel (forthcoming a) points out that
twenty-two out of fifty Rgvedic tribal names are not Indo-Aryan, with a majority of them
occurring in later books. He sees these as direct takeovers of local names of tribes or
individuals inhabiting the subcontinent before the arrival of the Indo-Aryans. While

98 The Quest for the Origin of Vedic Culture

this may well be the most economical explanation, one nonetheless needs to eliminate
the possibility, parallel to the one outlined throughout this section, that such tribes and
individuals may have been itinerant individuals or groups intruding upon a preexisting
Indo-Aryan community, as opposed to intruding Indo-Aryan-speaking groups intrud¬
ing upon non-Indo-Aryan ones. It seems relevant, in this regard, that Witzel (forthcom¬
ing a) notes that many of these names have not survived even in the Atharvaveda and
Yajurveda mantras, which could be taken to underscore their transience.

Place-Names and River Names

The non-Indo-Aryan nature of the terms and names noted earlier also has to be juxta¬
posed with the fact that the place-names and river names in northern India are almost
all Indo-Aryan. These names are, to my mind, the single most important element in
considering the existence of a non-Indo-Aryan substratum position. Unlike people, tribes,
material items, flora, and fauna, they cannot relocate or be introduced by trade (although
their names can be transferred by immigrants). In other words, it is difficult to exclude
the possibility that the foreign personal and material names in the Rgveda were intru¬
sive into a preexisting Indo-Aryan area as opposed to vice versa. This argument of lexi¬
cal transiency can much less readily be used in the matter of foreign place-names. Place-
names tend to be among the most conservative elements in a language. Moreover, it is
a widely attested fact that intruders into a geographic region often adopt the names of
rivers and places that are current among the peoples that preceded them. Even if some
such names are changed by the immigrants, some of the previous names are invariably
retained (e.g., the Mississippi river compared with the Hudson, Missouri state com¬
pared with New England).

In the 1950s Hans Krahe analyzed the river names in central Europe and found
them to be Indo-European. 34 This evidence has been used to argue that the homeland
must have been in central Europe, since had the Indo-Europeans intruded into this
area from elsewhere, they would have borrowed names from the local non-Indo-Euro-
pean groups that preceded them. More recendy, Theo Vennemann (1994) has argued
that these river names are actually not Indo-European at all, thereby suggesting that the
Indo-Europeans were intruders into the area after all, who adopted the local hydronymic
and toponymic forms, since “toponyms are rarely changed, they are merely adapted”
(264). In both cases, the assumption is drat place names are conservative.

With this in mind, it is significant that there are very few non-Indo-Aryan names for
rivers and places in the North of the Indian subcontinent, which is very unusual for migrants
intruding into an alien language-speaking area. Of course, it could be legitimately argued
that this is due to the Aryans’ Sanskritizing the names of places and rivers in the North¬
west (although this raises the issue of why the local flora and other names were was not
likewise Sanskritized). In the hydronomy of England, Celtic names are fewer in the east,
but they are preserved in major rivers (Hainsworth 1972, 45). On the other hand, they
become more frequent in the center, and more numerous in the west, a pattern that can
be correlated with the historical data on Saxon settlement, which would have been dens¬
est in the east, thereby explaining die fewer Celtic names in that area. Witzel finds the
same holds true in Nepal: “The whole west of the country has been Indo-Aryanized thor-

Linguistic Substrata in Sanskrit Texts 99

oughly and early enough ... as to eliminate most traces of earlier Tibeto-Burmese” (218).
Such non-Indo-Aryan names become more visible in other parts of the country. All this
is of comparative use to support the idea of Indo-Aryan migrations.

In terms of the oldest attested Indo-Aryan period, Witzel (1999) has done extensive
work on river names and place-names in the Rgveda, from which I will focus on the
Northwest. Witzel finds almost all the place-names in the Rgveda, which are few in
number, are Indo-Aryan, or at least Sanskritized. In his estimation, “most of the forms
are easily analysable new formations, so typical of settlement in a new territory” (368),
While this is a significant point, the lack of non-Indo-Aryan terms for toponomy and
hydronomy in this area immediately deprives us of essential data that have been funda¬
mental in establishing the existence of substrata in other languages. As a point of con¬
trast, classical Greek maintains only 40 names (from 140 toponyms in Homer) that are
Greek from the point of view of the classical language (but not necessarily Indo-Euro¬
pean, that is, they are new formations adopted after the break up of Proto-Indo-Euro¬
pean): barely one-third of the total. The remaining two-thirds are etymologically obscure
(Hainsworth 1972, 40). Such obscurity gives clear indication of a pre-Greek, non-Indo-
European substratum. The lack of foreign place-names in the oldest Indo-Aryan texts,
in contrast, is remarkable when compared with the durability of place designations else¬
where. The same applies to rivers. Witzel again notes that “such names tend to be very
archaic in many parts of the world and they often reflect the languages spoken before
the influx of later populations” (368-369). Yet here again, “by and large, only San-
skritic river names seem to survive” in the Northwest (370). In the Kuruksetra area, “all
names are unique and new formations, mostly of IA coinage” (377).

None of the river terms are Dravidian. The Ganges, which is the easternmost river
mentioned in the Rgveda, has an unusual etymology containing a reduplicated form of
gam ‘to go’, which Witzel believes is an old, non-Indo-Aryan loan, despite its Sanskritic
look. Later texts, however, mention rivers farther east and south from the Rgvedic home¬
land that show signs of Munda and Tibeto-Burmese influence in the northeast, and
Dravidian influence toward central India. For my purposes, paralleling the logic out¬
lined previously concerning the local fauna and flora, if the Indo-European had come
from the Northwest of India, one would also expect the terms for hydronomy and to¬
pography of this area to preserve acceptable Indo-European etymologies. In fact, the
hydronomic and topographic evidence is much more decisive than that of the flora and
fauna: river names and place-names cannot be loanwords or be the result of adstratum,
unlike the rest of the material overviewed in this chapter. If it can be convincingly dem¬
onstrated that the majority of Indo-Aryan hydronomic and topographic terms could not
have evolved from Proto-Indo-European, then the Indigenous case would lose cogency.
Of course, this procedure would be off to a rather tricky start by the very fact that we
can only guess at what any hypothetical Indo-European terms for places or rivers might
look like in the first place, but there is agreement about what is acceptable in terms of
Indo-European roots.

However, in “the ‘homeland’ of the Rgvedic Indians, the Northwest “we find “most
Rgvedic river names . . . are Indo-Aryan, with the possible exception of the Kubha,
Satrudri, and perhaps the Sindhu” (373). These latter, according to Witzel (1999),“prove
a local non-IA substrate. In view of the fact that Witzel has provided a list of thirty-
seven different Vedic river names, these two or three possible exceptions do not make

1 00 The Quest for the Origin of Vedic Culture

as strong a case as one might have hoped. All the rest can indeed be derived from Indo-
European roots. Morever, other scholars have even assigned Indo-Aryan etymologies to
two of these three possible exceptions. 35

Witzel’s reading (1999) of the evidence of hydronomy is as follows:

During the Vedic period, there has been an almost complete Indo-Aryanization of the
North Indian hydronomy. . . . Indo-Aryan influence, whether due to actual settlement,
cultural expansion, or. . . . the substitution of indigenous names by Sanskrit ones, was
from early on powerful enough to replace the local names, in spite of the well-known
conservatism of river names. The development is especially surprising in the area of the
Indus civilization. One would expect, just as in the Near East or in Europe, a survival of
older river names and adoption of them by the 1A newcomers upon entering the territo¬
ries of the people(s) of the Indus civilization and its successor cultures. (388-389).

Such conservatism is, indeed, extremely surprising, especially since the Indo-Aryans did
not enter in sufficient numbers to be perceivable in the skeletal record of the subconti¬
nent, as we shall see in the next chapters. One also wonders how such small numbers
of immigrants could have eradicated the names of rivers and places in the Northwest of
die subcontinent in die few hundred years that separated their arrival from the time the
texts were compiled—Witzel allows about seven hundred years from 1900 to 1200 b.c.e.
(74)—and yet not succeed in doing the same when they Aryanized the eastern and south-
central areas in the two or three millennia that followed (as I have noted, places in these
areas show signs of pre-Aryan indigenous Dravidan and Munda etymologies to this day).
Witzel agreees with Kuiper and others that the preexisting groups “must have had a
fairly low social position as they were not even able to maintain their local place and
river names, almost all of which were supplanted by new Sanskrit ones” (77). This
position needs to accommodate the fact that preceding the arrival of Indo-Aryans in the
Rgvedic homeland up to 1900 b.c.e. was the highly sophisticated urban culture of the
Indus Valley (which, as will be discussed in the next chapters, did not just evaporate
after the “decline” of the Indus Valley Civilization).

Place-names are not much more decisive in this matter either, although much more
work needs to be done in the area of present-day place-names in the subcontinent, which
have not received the attention they deserve. Growse (1883) was the first and one of the
only scholars to devote attention to this area, and his work is still a useful place to start.
From his perspective, “Neither from the intrinsic evidence of indigenous literature, nor
from the facts of recorded history, is it permissible to infer the simultaneous existence
in the country of an alien-speaking race at any period” (320). He has scant regard for
the etymologizing endeavors of those who attempted to identify a pre-Sanskritic-speak-
ing people on the basis of the place-names of North India: “The existence of such a race
is simply assumed by those who find it convenient to represent as non-aryan any forma¬
tion which their acquaintance with unwritten Aryan speech in its growth and decay is
too superficial to enable them at once to identify” (320). He further complained that “a
derivation from Sanskrit by the application of well-established but less popularly known
phonetic and grammatical laws, is stigmatized as pedantic” (320).

Growse found that place-names in the North consisted of those compounded with
an affix denoting place; those compounded with an affix denoting possession; and those
without an affix, being an epithet of the founder or of some descriptive feature of the
place. He found the most numerous were those in die first category compounded with

Linguistic Substrata in Sanskrit Texts 101

the affix pur ‘city, urban center’ (discussed more fully in the next chapter)- which trans¬
formed into a number of forms such as - oli,-uri,-uru. He quotes a verse from Vararuchi’s
Prakrit grammar to show how inital p-, among other initials, can be elided. 36 He contin¬
ued to use Prakrit rules to determine a number of other affixes and concluded his study
with a statement with which many would still agree: “So many names that at a hasty
glance appear utterly unmeaning can be traced back to original Sanskrit forms as to
raise a presumption that the remainder, though more effectively disguised, will ultimately
be found capable of similar treatment: a strong argument being thus afforded against
those scholars who hold that the modern vernacular is impregnated with a very large
non-Aryan element” (Growse 1883, 353).

In contrast, the only other recent study of place-names in the North of which I am
aware, 37 which claims to be comprehensive, is dedicated to demonstrating that “before
the Aryan invasion India was inhabited by the ancestors of the Dravidians, who mi¬
grated from the mediterranean region” (Das and Das 1987, 2). Even with such a start¬
ing premise, the book scarcely produces half a dozen possible Dravidian names from
the Northwest of the subcontinent: the vast majority are from the east—Bengal and Uttar
Pradesh—which few would dispute were originally settled by non-Aryans. 38 Southworth
(1995) has argued for the existence of Dravidian place-names in Maharashtra, Gujarat,
and Sindh. My concern is with the Northwest of the subcontinent. In this regard,
Southworth has made some tentative identifications of a few place-names in Sindh ending
in -uiari, wari, and the Punjab - wall. 39 As I have noted earlier, Growse (who was exam¬
ining names in the Uttar-Pradesh area) considered similar forms to be Prakritizations of
pur and therefore Indo-Aryan. Mehendale considers the 2,045 wadi settlements in the
Retnagiri area that he surveyed to have come from the Sanskrit form vdtika (a possible
Prakritization of Sanskrit vrt). In short, there is much to be done on the subject of place-
names in the Northwest.

Apart from these observations, Witzel (forthcoming, 12) notes three place-names in
Kashmir ending in -musa and a river called Ledari that he considers non-Indo-Aryan. In
any event, he concludes: “In light of the present discussion about the arrival of the
Aryans in India and in some circles of Anglophone archaeology, that is the growing
denial of any immigration or even trickling in of people speaking Indo-Iranian or Indo-
Aryan dialects, it is important to note that not only the Vedic language, but the whole
complex material and spiritual culture has somehow been taken over and absorbed in
the northwest of the subcontinent” (72).

Of substantial importance is Witzel’s discovery (forthcoming b) that there was no
Dravidian influence in the early Rgveda. He divides the Rgveda corpus into three dis¬
tinct chronological layers on linguistic grounds and finds that Dravidian loans surface
only in layer II and III, and not in the earliest level at all: “Consequently, all linguistic
and cultural deliberations based on the early presence of the Drav. in the area of speakers of
IA, are void” (17; italics in original). Instead, “we find more than one hundred words
from an unknown prefixing language” that is neither Dravidian, Burushaski, nor Tibeto-
Burmese (6). On the basis of certain linguistic evidences, such as Munda-type prefixes
(ka-, k i-, fct-, ku-, fee-, and “double prefixes”), he prefers to consider the pre-Aryan lan¬
guage an early form of Munda. 40 He finds the same prefixes in later texts whose geo¬
graphic boundaries are farther east. These deductions, combined with the known posi¬
tion of Munda in the east, cause Witzel to postulate a Munda substramm in the oldest

102 The Quest for the Origin of Vedic Culture

Rgvedic period. He considers that an essential corollary of his findings is “that the lan¬
guage of the Indus people, at least those in the Punjab, must have been (Para-) Munda or
a western form of Austro-Asiatic (12). He proposed that this language, in turn, was an
overlay over an unknown, lost language (Masica’s “language x”).

As a side note, Witzel draws attention to another interesting point. Since the Indus
Valley was a trading civilization, why did non-Indo-Aryan terms for trade not surface in
the Rgveda? Why are all the loans identified primarily terms for fauna, flora, and agri¬
culture (mostly from “language x”)? Since most migrationists would accept that the Indo-
Aryans interacted with the tail end of the Indus Valley Civilization, one would have
grounds to expect that incoming Indo-Aryans would have borrowed trading terms from
this civilization, which is not the case.

Dravidian, in Witzel’s scenario (forthcoming b, 30), was a later intruder that, inter¬
estingly, he is prepared to consider as having arrived at about the same time as the
Indo-Aryan languages, (30), explaining the subsequent influences of Dravidian on later
Vedic [strata] (Dravidian, in the process, also absorbed retroflexes and lexical items for
flora, and so on, from the unknown, preexisting language). 41 This causes me to again
raise the previous consideration—perhaps an inescapable one from the perspective of
Indigenous Aryanism—and one that has yet to receive scholarly attention. What is the
possibility of all the various innovations noted here being the result of alien languages,
whether Dravidian, Munda, or anything else, intruding on an Indigenous Indo-Aryan
language as opposed to vice versa? Or of adstratum or superstratum relationships as
opposed to substratum? Witzel has provided data to argue that this certainly must have
been the case with Dravidian, since Dravidian influence is not visible in the earliest
layers of the Rg but only in subsequent layers: “Such words could have been taken over
any time between the RV . . . and the earliest attestation of Tamil at the beginning of
our era” (31). He notes that most of the eight hundred words assigned a Dravidian
etymology by the Dravidian Etymological Dictionnary are attested only in the later Ep¬
ics or classical Sanskrit texts: “The Indo-Aryans did not at once get into contact with
speakers of Drav. but only much later, when the tribes speaking IA were already living
in the Panjab and on the SarasvatT and Yamuna” (19). If Dravidian has influenced Indo-
Aryan through adstratum or superstratum interactions, and not through a substratum
relationship, then why could Munda (or other languages) not likewise have done so?

Indo-Aryan, or Dravidian and Munda Migrations?

I wish to further explore this possibility, first raised by Bloch, namely, that it was Dravidian
that intruded into an Indo-Aryan speaking area and not vice versa. It seems to me that
an Indigenous Aryan position would be forced to consider this possibility in one way
or another. Linguistically, at least, there does not seem to be any reason that this could
not have been the case. Brahui, in this scenario, could have been a language pocket of
Dravidian that remained stranded in the North after the rest of the Dravidian speakers
had continued down south. This would fit with the claim that Brahui is better con¬
nected to the southern language group of Dravidian rather than the northern one. While
I will concentrate on Dravidian, here, since this has been the focus of most research in
this area, the logic being outlined is just as applicable to Munda (which, until the work

Linguistic Substrata in Sanskrit Texts 103

of Kuiper and Witzel, has received less scholarly attention in terms of its influence on
Vedic than Dravidian), and even “language x”.

There is also another very significant reason that Indigenous Aryanists would have
to argue for a post-Indo-Aryan arrival of Dravidian (Witzel, personal communication).
If the innovative features Indo-Aryan shared with Dravidian and/or Munda and/or other
unknown languages, such as the retroflexes, were the result of adstratum influences
between these languages in the proto-historic period, as has been presented as a possi¬
bility earlier, one would expect that some of these linguistic features would have rippled
out into other adjacent Indo-European areas, or at least into neighboring Iranian. After
all, from the perspective of a South Asian homeland, the Iranians could not have left
the subcontinent much before the composition of the Rgveda due to the similarity of
the languages. Therefore, from the perspective of a language continuum homeland with
the Northwest of the subcontinent as its nucleus, Iranian would have been the closest
to this nucleus. Why, then, did Iranian not share these innovations? Wiry do most of
these South Asian areal features seem to stop at the Khyber Pass, so to speak?

In this regard, Hock (1993) notes that some of the innovations are actually shared
by eastern Iranian, specifically retroflection: “The core area of the change must have
been in South Asia proper, from which the change spread only incompletely to the
Nuristani and East Iranian languages on the northwestern periphery, before coming to
a complete halt in geographically even more remote Iranian territory” (96). He has also
argued that a second of the three main features discussed in this chapter, namely, the
postposed quotative marker iti, could have been paralleled by Avestan uitl (although
Kuiper [1991] feels this lacks any foundation). There have also been claims of loans
from Dravidian into Avestan as well as A’edic, which have been construed as coming
from the Indo-lranian period (see Southworth 1990 for examples). Nonetheless, from
the Indigenous Aryan perspective, it would be easier to argue that Dravidian and/or
Munda, and/or “language x” speakers, intruded into an Indo-Aryan-speaking area after
Iranian had already left, and that consequently the innovations were the result of a
superstratum of these language speakers settling in Indo-Aryan-speaking areas in the
Northwest.

Alternatively, these languages could have skirted the Indo-Aryan languages in the
Northwest and influenced them as adstratum. The first issue to be dealt with in this
case, of course, is chronological. Either it would have to be argued that the Dravidian/
Munda/“language x” speakers entered die subcontinent after about 1900 b.c.e. and
interacted with the Indo-Aryans as adstratum or superstratum during the period prior
to the commonly accepted composition of the hymns. If this is too late for proto-Dravidian,
it would have to be argued that the Indus Valley was Indo-Aryan and that the Vedic
texts are far older than philologers have so far dated them. Both these latter issues will
be discussed at length in the following chapters.

In terms of the possibility of a later Dravidian intrusion into the subcontinent, I will
briefly review some of the theories pertaining to the chronology and origins of Dravidian.
There is no consensus regarding the origin of Dravidian. McAlpin’s attempt (1974) to
connect Dravidian with Elamite is the most often quoted endeavor, although often without
much critical analysis of the claims involved (perhaps because so few linguists are com¬
petent enough in the two languages involved to evaluate his work). The Dravidian lin¬
guist Krishnamurti (1985) appears unconvinced by this idea and wonders if “McAlpin

104 The Quest for the Origin of Vedic Culture

was carried away by the flight of his imagination.” As far as he is concerned,“many of
the rules formulated by McAlpin lack intrinsic phonetic/phonological motivation and
appear ad hoc, invented to fit the proposed correspondences,” and “McAlpin’s foray
into comparative morphology is even more disastrous” (225-226). Other reactions to
McAlpin’s proposal, by Emeneau, Jacobsen, Kuiper, Rainer, Stopa, Vallat, and Wescott,
have also been very reserved (although Paper and Zvebevil were more enthusiastic). (These
responses were published in Current Antropology volume 16 [1975].)

Vacek (1987) gives an overview of scholarship attempting to connect Dravidian with
the Altaic languages. While he is partial to this position, he is forced to acknowledge
that “conclusions concerning the type of their relations are premature because the avail¬
able data can often be interpreted in various ways” (1 2). Sjoberg (1971) also undertakes
such an overview but is more partial to the attempts at finding connections with the
Uralic languages, although admiring that this is at odds with genetic data pointing to
links with southwestern Asia (16). Uralic was also favored by Tyler (1968). Another
attempt at establishing genetic relationships has been with Japanese. 42 A further group
of scholars see the typological or other features illustrated by all these efforts as evidence
of a superfamily, Nostratic. Nostratic is a term coined by the Russian linguist Illich-
Svitych to refer to a superlanguage family, or a protofamily of protofamilies. Depending
on the linguist, this might include Afro-Asiatic, Elamite, Kartvelian, Uralic, Altaic, and
Dravidian in addition to Indo-European, although many linguists believe this language
is completely beyond the ability of current techniques in linguistics to demonstrate.

In any event, clearly, the origins of Dravidian are yet to be established; as Sjoberg
(1971) concludes, “we can only speculate as to the time and place of the initial forma¬
tion of a distinctive Dravidian people and culture” (17). Less work has been focused on
the origins of Munda on the subcontinent, but here, too, any dating attempt can only
be highly speculative. D’iakonov (1997) tentatively explores the possibility of its con¬
nection with Sumerian. Acknowledging that there are no “amazing similarities,” he
nonetheless hopes that “some suggestive material may perhaps emerge” (58). As for
“language x,” since it is primarily a hypothetical language, there are no grounds whatso¬
ever for determining its chronology or point of origin (unless Kuiper’s linguistic pat¬
terns can be correlated with other language families).

Chronologically, scholars have little of substance upon which to base their dates for
the incursion of Dravidian into the subcontinent (all do seem to agree that Dravidian
was not an indigenous language). Zvelebil (1972) considers them “a highlander folk, 43
sitting, sometime round 4000 B.C., in the rugged mountainous areas of Northeastern
Iran. . . whence, round 3500 B.C., they began a Southeastern movement into the In¬
dian subcontinent which went on for about two and a half millennia” (58). Needless to
say, since the Elamite connection has not been widely accepted, there are no obvious
traces of Dravidian outside the subcontinent that can determine either its point of ori¬
gin, its chronology, or the overland route of its speakers, although attempts have been
made to find traces of them in central Asia (e.g., Lahovary 1963). Pejros and Shnirelman
(1989) volunteer a date of 3000 b.c.e. for proto-Dravidian without stating their grounds
and hold that the language must have entered the subcontinent from the Northwest
due to its Nostratic connection.

In reality, any attempt to establish a date for proto-Dravidian is ultimately, as Zvebevil
(1972) acknowledges, “in the nature of guesswork,” since glottochronology, as I will

Linguistic Substrata in Sanskrit Texts

105

discuss in chapter 13, has been almost unanimously discredited. As far as I can deter-
mine, there is very little that is decisive that can be brought forward to deny the possi¬
bility that Dravidian or Munda speakers intruded upon an Indo-Aryan speaking area
and not vice versa. This possibility would be reinforced if the claims of a greater antiq¬
uity for the Indo-Aryan language could be established. We might also bear in mind
Bloch’s suggestion that such intrusions could have been the result of individuals as
opposed to major population movements, just as individual Sanskrit speakers coming
from North India massively affected the southern languages (which became heavily
Sanskritized) without migrating down in vast numbers.

Conclusion

There might be scope, in all this, for considering alternative models to that of invading
Aryans borrowing a specialized lexicon from Dravidian, Munda, or linguistically un¬
known indigenous people. Indigenist suspicions are initially aroused due to the consid¬
erable differences in the opinions of the foremost authorities in this area. Some schol¬
ars are quite prepared to acknowledge the inconclusiveness of the linguistic evidence.
Other linguists have concluded, both because of the syntactical reasons discussed ear¬
lier and because Dravidian and Munda can account for only a small minority of the
unaccountable words, that unknown, extinct languages must have existed in the
protoperiod. This is by no means an unreasonable proposal. There are a number of
languages on the subcontinent apart from Dravidian and Munda. Tibeto-Burmese is
the most widespread, but there are also the language isolates such as Kusunda, Nahali,
and Burushaski that have been examined by Witzel (1999) as possible substratum can¬
didates. Burushaski is of particular relevance, since it is situated in the Northwest.
However, neither this nor any other known language has been recognized by special¬
ists as a possible candidate for the innovations in the Rgveda. 44 Hence the need for
“language x.”

The problem is that the existence of such possible extinct languages is very hard to
verily; Kuiper’s attempt at pinpointing consistent alien structural patterns in Indo-Aryan
might be the nearest one can hope for in terms of “proof.” Emeneau (1980), a propo¬
nent of a Dravidian substratum, seems to recognize that resorting to such opaque expla¬
nations as extinct languages is hardly likely to satisfy empirically minded historians: “It
hardly seems useful to take into account the possibility of another language, or language
family, totally lost to the record, as the source [of the foreign words]” (169). Resorting
to such explanations is seen as rather desperate pleading by the frustrated Talageri (1993),
who “cannot proceed with these scholars into the twilight zone of purely hypothetical
non-existent languages” (200; italics in original). 45 Mallory (1975) opines that “the reli¬
ance on simple a posteriori appeals to unknown (and perhaps non-existent) substrates
to explain linguistic change should be dismissed from any solution to the IE homeland
problem” (160). As has been noted, such a hypothesis can be neither verified nor falsi¬
fied and thus is incapable of resolving this debate. Moreover, even if it could be veri¬
fied—and, in deference to Kuiper’s work, unknown languages can be “proved” if consis¬
tent phonemic or morphological patterns can be identified in textual sources—how can
one discount the possibility that such linguistic influences could still be explainable along

106 The Quest for the Origin of Vedic Culture

the same lines suggested previously: as resulting from adstratum, as opposed to substra¬
tum, relationships?

In summary, all these linguists are operating on the assumption, based primarily on
other criteria, that the Aryans “must have” invaded India, where there could not have
been a “linguistic vacuum.” All alien linguistic features identified in Sanskrit texts have
accordingly been explained as belonging to pre-Indo-Aryan substrata. Since Dravidian
and Munda inadequately explain these changes visible in Sanskrit, many are forced to
consider theories of extinct languages. How the data could be convincingly reinterpreted
if this assumption were to be reconsidered remains to be seen, since a comprehensive
and objective case has yet to be made by the “Indigenous Aryanists” despite their possess¬
ing the rudiments of a variety of alternative explanations already advanced by Western
linguists. As I have attempted to outline, loanwords can enter a language in many ways
without the need for postulating a substratum (or even an adstratum). Many of the for¬
eign terms for flora and fauna could simply indicate that these items have continually
been imported into the subcontinent over the centuries, as continues to be the case
today. The exception to this is place-names and river names, but the absence of foreign
terms for the topography and hydronomy of the Northwest deprives us of significant
evidence that has been used to establish substrata elsewhere. Most important, the pos¬
sibility of spontaneous development for many of the syntactical features common to
Sanskrit and Dravidian and Munda, coupled with the possibility of an adstratum rela¬
tionship for features that are undoubtedly borrowings between the languages, are the
most obvious alternative possibilities that need to be fully explored.

Thomason and Kaufman (1988) have outlined a typology of change typically caused by
the cultural pressure of a language on another—the more overpowering the influence, the
more the language will transform. Casual pressure results in lexical borrowings only; less
casual influence produces lexicon and minor structural borrowing; more intense contact
increases the amount of structural borrowings; and strong and very strong cultural pres¬
sure result in moderate and extensive structural transformation. I am not aware of any
technique available to present linguistic knowledge that, in a protohistoric setting, can
determine whether such influences between languages—whether they be lexical or syntac¬
tical—are the result of adstratum, substratum, or even superstratum relationships.

Salmons (1992) notes the same concerns in his search for substratum influence in
Northwest Indo-European vocabulary: “Adstratal borrowing or even internal innova¬
tions, not just substratal borrowings, might show these previously prohibited forms.
Again, simple alternative explanations to the substrate hypothesis seem to present them¬
selves” (274). He goes on to state that “as a result of the proclivity to speculate, sub¬
strate explanations carry a bad reputation among historical linguists. ... all other av¬
enues must be exhausted before we reach for a substrate explanation” (266). Caution
must be exercised that substratum explanations are not resorted to as a kind of conve¬
nient linguistic dumping ground where anything that does not fit into the dominant
recorded culture is heaped by default.

In conclusion, the theory of Aryan migrations into the Indian subcontinent would
better be established without doubt on other grounds, for research into pre-Aryan linguis¬
tic substrata to become fully conclusive. That Indo-Aryan intruded onto a non-Indo-
Aryan substratum still has much to recommend itself. Perhaps it is the least compli¬
cated way of accounting for the available evidence, but it is not without limitations. To

Linguistic Substrata in Sanskrit Texts 107

reiterate, the main alternative possibilities that Migrationists need to eliminate are: (1)
that Indo-Aryan could have spontaneously originated some of the non-Indo-European
innovations visible in it and then shared these with Dravidian and/or Munda (or vice
versa); (2) that the non-Indo-European words in Sanskrit texts from Dravidian, Munda
and/or “language x” are simply loans resulting from trade or other nonsubstrarum in¬
teractions between language groups; and (3) that any alien linguistic features in Sanskrit
texts that cannot be accounted for by possibilities 1 or 2, whether phonemic or mor¬
phological, could be the result of adstratum (or superstratum) rather than substratum
contacts. The other possibility that needs to be eliminated is that the Indo-Aryan names
of places and rivers could not have evolved from Proto-Indo-European by inherent and
natural internal linguistic developments. As for Indigenists, they must accept that any
discussion of Indo-Aryan origins that neglects the substratum data, simply cannot be
taken seriously.

All this resonates with Polome’s conclusion (1990b) to his researches on Germanic
substratum: “ In many cases the evidence remains inconclusive, and only when extralinguistic
evidence can be coordinated with the lexical data can we posit a ‘substrate’ origin of the
terms” (285). In short, while certainly suggestive, it is difficult to see how the “evidence”
of a linguistic substratum in Indo-Aryan, in and of itself, can be used as a final arbitrator
in the debate over Indo-Aryan origins.

Linguistic Paleontology

As mentioned previously, linguistic paleontology was inaugurated by Adolphe Pictet in
1859 in three volumes that covered every imaginable set of Indo-European cognates. 1
This method was fundamental in relocating the Indo-European homeland away from
the East, where the early scholars had preferred to situate it. Just as paleontology in¬
volves attempting to understand the plant and animal life of previous geological ages
from fossils found in the archaeological record, linguistic paleontology involves hypoth¬
esizing about the social, religious, political, economic, ecological, cultural, and geographic
environment of protohistoric cultures from linguistic fossils, or cognate terms, preserved
in the various members of a language family. As Otto Schrader (1890) put it, “As the
archaeologist. . . descends into the depths of the earth ... to trace the past in bone and
stone remains, so the student of language might . . . employ the flotsam and jetsam of
language ... to reconstruct the picture of the primal world” (iii). Once a picture embel¬
lished with details such as flora, fauna, landscape, and economy has been formed by
this method, the idea is to attempt to situate it in an appropriate geographic setting in
the real world and then connect it with a corresponding archaeological culture, Nietzsche
was to compare the philologist to an artist touching up an old canvas. In this case,
however, the canvas was well over five thousand years old. Could philology bring this
completely faded picture back to life, or would it paint right over it and create a com¬
pletely different landscape?

This section will outline some of the features of this method that have been relevant
to the history of the quest for the Indo-Aryans, or that have attracted responses from
the Indigenous school. Since I am not an Indo-Europeanist myself and my audience is
primarily scholars of South Asia who are interested in the protohistory of the Indian
subcontinent, I will not attempt to represent most of the discussion and debate amongst

108

Linguistic Paleontology 109

linguists concerning technical details such as the protoforms of words diat are relevant
to this section, but will address the more general conclusions drawn from them. I should
also note that most present-day Indo-Europeanists are fully aware of the limitations of
this method, and of its checkered history. However, much that will be considered passe
to specialists in the field still surfaces in books on Indian proto-history and therefore
remains relevant to the purposes of this work.

Flora and Fauna

One set of cognates, which became extremely influential in supporting a German home¬
land, involved the term for the common beech tree. As Friedrich (1970) notes, “The
botanical beech line, partly because it has been so often misused, has guaranteed this
tree a sure place in all discussions of the Proto-Indo-European homeland” (106). Pictet
triggered the popularity of this tree among homeland-seekers by presenting an array of
cognates for this term from all the Indo-European languages accessible in his time. Since
this tree had cognates in both the Indo-Iranian and the European side of the family, it
was assumed to have existed in the proto-language before the various linguistic branches
separated. Words with cognates in only the western (or only the eastern) branches retain
the possibility that their referents might have been encountered after the common Proto-
Indo-European period in a secondary, western (or eastern) location, and therefore not
indicative of the original homeland. As mentioned previously, scholars such as Geiger
used this information to draw up maps of the geographic boundaries within which the
beech tree grows—specifically, German-centered Europe (thus excluding the Asian
hypothesis that was still almost universally accepted in Pictet’s time)—and the Aryan
homeland was set within this area. The beech evidence was particularly used by Thieme
(1964, 597) who argued for a homeland between the Black Sea and the Baltic with an
eastern border fixed by the boundaries of the beech habitat.

There are various problems with this approach. It has been noted that the beech is
linguistically unattested in Anatolian, but this language was spoken in the very area where
scholars believe the beech was native. In odter languages spoken in the heartland of beech
territory, the word was transferred to refer to the ‘oak’ so dtat “the concatenation of as¬
sumptions required to press the ‘beech line’ into argument would appear to be exceed¬
ingly dubious” (Mallory and Adams 1997, 60). Friedrich (1970) points out further limi¬
tations of the birch evidence based on its shifting habitat and concludes that “none of
Thieme’s well-known criteria support his homeland hypothesis” (30). The area where the
beech, or any tree, grows now may not be the same as where they might have grown many
millennia ago. Paleobotony might help locate prehistoric trees to some extent, but Friedrich
explicitly encourages the philologist to “retain a due skepticism of ‘hard science’” (14) 2

But such methods have other limitations when it comes to locating homelands.
Friedrich (1970), in his taxonomy of Indo-European trees, proposes that linguistic
paleontology reveals eighteen categories of trees that were known to the ancient Indo-
Europeans. 3 His findings reveal that all three divisions of the Slavic languages have at
least one of the reflexes for each of these eighteen terms, indicating that the Slavs
were familiar with all eighteen Indo-European trees; the correspondence, in this case, is
100 percent. This suggests, to him, that the speakers of the common Slavic period lived

110 The Quest for the Origin of Vedic Culture

in an ecological, that is, arboreal, zone similar or identical to that of the Proto-Indo-
European’s (167). 4 In sharp contradistinction, the paucity of these eighteen stocks attested
in Indie, Anatolian, and Tocharian suggests to Friedrich “substratum influence” or
“movement into a radically different environment” (169). This evidence is taken as sig¬
nificant evidence that the Indo-Europeans must have come into India from elsewhere
(e.g., Possehl 1996, 65). Friedrich himself seems aware of the possible objections that
drawing too far-reaching conclusions from his results might provoke and is hasty to add
that he “would be the first to insist that the arboreal evidence cannot be used in isola¬
tion to construe a conclusive argument for a Proto-Indo-European homeland in the
Ukraine or the Cossack steppe” (168).

The immediate objection from the perspective of the Indigenous Aryan school was
first articulated by Dhar, head of the Sanskrit department at Delhi University in 1930,
when confronted with similar arguments. Dhar’s (1930) is the first serious attempt that
I can trace to challenge the prevailing ideas regarding the Aryan invasions on linguistic,
as opposed to philological, grounds: 5

Central Asia might be the secondary home of the Aryas (Indo-European’s] . . . but their
primary home might be situated outside central Asia, in the Himalayas. ... Of late, the
beech argument is much advertised by the promoters of the Indo-European theory of the
home of the Aryas. But the term for the “beech” might have been coined by the Aryan
settlers in Europe only where the tree grew. (26)

Tire logic here is that if the Indo-European tribes had, hypothetically, journeyed forth
from an Indian homeland, they would obviously have encountered strange trees, animals,
and fauna that did not exist on the subcontinent and for which they would have coined
new terms or borrowed names from the indigenous people resident in those areas. Sub¬
sequent Indo-European tribes would have adopted the same terms from their predecessor
Indo-Europeans resident in the places where the unfamiliar items were encountered. 6 Such
new lexical terms would obviously not surface in the Indo-Aryan languages that remained
behind in the subcontinent, since the objects they denoted did not exist in India. Nor
would they surface in other Indo-European languages such as Tocharian and Anatolian,
which were geographically removed from the well-trodden northwestern path taken by
most of the Indo-European tribes that eventually resurfaced in the west. The result would
be a large number of common terms in the western Indo-European languages (since they
are numerically greater) and a smaller number in lndo-Iranian. 7

Friedrich’s results, then, indicating a paucity of his reconstructed tree stocks in Indie,
Anatolian, and Tocharian, would not be incongruous to the Indigenous Aryan posi¬
tion. On the contrary, anyone postulating a South Asian homeland would anticipate
such findings. Dhar’s basic premise can be used to challenge conclusions drawn from
any other cognate terms of material culture extant in the western Indo-European branches
but absent in the lndo-Iranian or Indo-Aryan ones. As Polome (1990b, 274) notes, there
are two equally logical ways of accounting for the lacuna of some linguistic feature in
one particular language that is shared by its cognates in other languages: either it was
never there to start with or it has been lost somewhere along the way. Indeed, other
linguists use exactly the same arguments as Dhar has used to account for items recon¬
structed in Proto-Indo-European that happen to be absent in their proposed homeland:
“Part of these terms cannot be reconstructed for the period of proto-Indo-European unity,

Linguistic Paleontology 111

but only for later dialect groupings. . . . hence the picture of the ancient Indo-Europeans’
plant and animal world is to be thought of as . . . one which changed as speakers of the
dialects moved to their later territories” (Gamkrelidze and Ivanov, 1995, 573); “What is
especially interesting about these words is that most of them denote natural objects . . .
typical of Europe and less typical of SW Asia. My impression is that these words were
borrowed when the W [west] Indo-European ethnic community migrated from some
region of SW Asia ... to Europe . . . and got acquainted with objects of nature which
had been unfamiliar or less familiar earlier” (Dolgopolsky 1989, 18). 8

Along very similar lines, another group of cognate terms was prominent in attempts
to locate the homeland in Europe or southern Russia. Thieme in particular held that
the term for ‘salmon’ is “especially characteristic” of the Indo-Europeans. According to
him, this fish is found only in the rivers that go into the Baltic and German Seas (1964,
597). 9 The salmon evidence is still in circulation, especially among those promoting a
northern German homeland (e.g., Diebold 1991, 1 3). 10 However, the salmon case is
slightly more complicated for the Indigenous Aryan school, since, in this case, Sanskrit
might have a cognate term ( laksa , ‘lac’) with the same etymology that has been assigned
to the Proto-Indo-European form for the salmon (*loks). If the Sanskrit form is, indeed,
a cognate, then how did the word come to mean ‘lac’ in Sanskrit, and ‘salmon’, or
‘fish’, in other languages?

Since Sanskrit also has a term laksa, which means a very large number, Elst (1996),
who argues for a South Asian homeland, has proposed that Indo-European tribes, upon
leaving the subcontinent, came across unfamiliar fish in large shoals to which they gave
the term 'numerous; hundreds of thousands’. 11 The interchange of number terms with
the nomenclatures of species that cluster together in multitudes is not uncommon. Elst
compares the Idksa/laksa case with the Chinese use of an insect character, wan, to de¬
note ten thousand. This general term for fish, which was preserved in Tocharian, then
eventually entered into some Indo-European languages to refer to more specific types of
fish. The word was applied to ‘salmon’ (Old High German lahs, Russian losos, etc.)
when the speakers of these languages encountered this specific reddish species of fish
(perhaps prompted by the almost identical Indo-European word for red) and, in other
languages, such as in Iranian Ossetic, to trout ( lasag ). 12 In any event, some linguists
claim that the Indie forms (particularly laksa) are not actual cognates, in which case it
could be argued that the word for ‘salmon’ could have been coined by tribes after they
had left the subcontinent along the lines outlined earlier. Using similar arguments,
Gamkrelidze and Ivanov (1983b) state that “in the specific meaning of‘salmon’ . . . the
word is common to the ‘Ancient European’ dialects and Eastern Iranian. . . . Of course
the word would have acquired this meaning in those areas where salmon was found, in
regions near the Aral or Caspian Seas” (77). 13

These are the types of arguments that have to be made to account for any terms
either not preserved in Indo-Aryan but present in other Indo-European languages or
preserved in Indo-Aryan but with a different meaning from cognates in other Indo-
European languages. 14 What must be noted is that scholars, such as Thieme, have used
exactly the same series of deductions, but in reverse. Indeed, as with so much of this
debate, Elst has basically redirected Thieme’s exact arguments. In Thieme’s scenarios,
(1953, 552), a protoword for ‘salmon’ in a salmon-breeding homeland gets transferred
onto other fish by tribes moving out of the salmon area and becomes a number, or

112 The Quest for the Origin of Vedic Culture

adjective meaning red, in India. Both ways of presenting the series of semantic trans¬
feral are, arguably, possibilities. 15 Clearly, in all homeland explanations, a certain amount
of juggling has to be done to account for all the available data, but we find other lin¬
guists such as Gamkrelidze and Ivanov utilizing similar arguments to Dhar and Elst in
defense of their Near Eastern homeland. 16 This brings us back to the focus of this in¬
quiry. Can India be convincingly denied the status of an urheimat contender by the
method of linguistic paleontology?

India has repeatedly been excluded as a potential homeland based either on the logic
outlined here, that is, that it does not have cognate forms for items of material culture
attested in other languages, or, by the inverse logic, that exotic items unique to India are
unattested elsewhere. This latter process of elimination has been consistently used to
exclude South Asia. Thus Thieme (1964) notes: “We can eliminate [as homeland can¬
didates] those [languages] for whose characteristic plants and animals no reconstructable
designations are available, that is India: (no Proto-Indo-European words for elephant,
tiger, monkey, fig, tree, etc.), [and] Iran (no proto-Indo-European words for camel, donkey,
lion, etc.)” (596; see also Bender, 1922, 21). More recently, Witzel (1995a) has remarked
along the same lines that “turning to Sanskrit, it is interesting . . . that ‘tropical’ [Indo-
European] words are . . . absent in it, which indicates that it was an immigrant into
South Asia. Words for lion, tiger, elephant are either loanwords from local languages, or
are new formations, such as hastin ‘elephant; the one with a hand’” (101). This argu¬
ment basically holds that since the terms for exotica typical of India have no cognates
elsewhere, these terms could not have been in Proto-Indo-European, and therefore Proto-
Indo-European could not have been spoken in the areas, such as India, where such
exotica are to be found.

Similar arguments were actually countered over a century ago by Western scholars them¬
selves. Lassen ([1851] 1867), as mentioned in chapter 1, was the first to attempt to deny
India the possibility of being the homeland on these very grounds that the other Indo-
European languages lacked terms for the exotica present in India. His reasoning was im¬
mediately dismissed by his colleagues: “The want of animals specifically Asiatic . . . can be
explained simply by die fact of these animals not existing in Europe, which occasioned
their names to be forgotten” (Weber 1857, 10). Max Muller ([1857] 1985) also rejected
this line of argument: “And suppose that the elephant and the camel had really been known
... by the united Aryans, when living in Asia, would it not have been natural that, when
transplanted to the northern regions, dieir children who had never seen a camel or ele¬
phant should have lost the name of them?” (101). Keith (1933) likewise complained:

Nothing is more unsatisfactory than to attempt to define Indo-European society on the
assumption that the Indo-Europeans knew only what can be ascribed to them on con¬
clusive evidence. Ex hypothesi, there were great dispersals of peoples from the original
home, and those who wondered away were unquestionably constantly intermingling
with other peoples . . . and it is not to be wondered at that in new surroundings new
words were employed; still less can it be a matter of surprise that peoples which ceased
to be in contact with natural features soon dropped the names which had become use¬
less. (189-190)

Lassen’s reasoning occurs in the Cambridge History of India, wherein Giles (1922)
had stated that “the primitive habitat from which the speakers of these languages de¬
rived their origin ... is not likely to be India, as some of the earlier investigators as-

Linguistic Paleontology 113

sumed, for neither the flora nor fauna, as determined by their language, is characteristic
of this area” (68). Dhar (1930) again rose to the challenge:

The absence of common names in the Indo-European languages for such Asiatic animals
as the lion and the tiger and the camel, cannot prove the European origin of the Aryas
[Indo-Europeans], for the names of such animals as are peculiar to the East might easily
be forgotten by the people [after they had left India) in die West where those animals
were not found. Or it is very probable that there may be several synonyms for the same
object in the Aryan mother tongue—the one tribe of the Aryas in Asia or India having
taken the fancy for one name while the other for another. . . . Professor Giles is an advo¬
cate of the European home of the Aryas. He ought to realize that his argument cuts both
ways, for the names of European flora and fauna do not exist in the Asiatic Aryan lan¬
guages either. Really it should not be difficult to understand that the names for trees and
animals disappear as the trees and animals themselves disappear. (30)

Dhar’s reasoning is simple but logical, and it returns to the same basic point. If the
Indo-European’s had migrated from India, it would, indeed, be possible that the words
for uniquely Indian objects would disappear from use and would not surface in the
western cognate languages. This exactly mirrors the logic outlined in the previous sec¬
tion in reverse: the newly coined words in the western languages to describe exotic items
not extant in India would obviously not be evidenced in the Indo-Aryan languages re¬
maining in India.

Here, again, we find present-day Western scholars reiterating exactly die same argu¬
ments: the importance of terms in the protolanguage designating plants, animals, and
other geographically bound concepts

should not be overestimated. If a given proto-language was spoken in an area outside that
of its daughter languages, specific words designating features of the ancient habitat are
not usually preserved in the attested languages. Therefore, if a language ancestral to a
group of European languages originated in Africa, we would not be able to find in the
extant lexical stock ancient words for “giraffe” and “elephant” which could suggest its
African origin. (Dolgopolsky 1987, 8)

Regarding the possibility of synonyms, Polome, (1990b) along the same lines as Dhar,
also objects to speculations that “fail to take into account such basic facts as the possi¬
bility that several designations . . . [for words in Indo-European] may have coexisted,
differentiated by the context in which they appeared and the people who used them”
(270). Furthermore, as Dhar notes, the argument cuts both ways: why should the Indie
languages be held accountable for containing the names of exotica not evidenced in the
western languages, and the western languages not have to account for their unique terms
with no Indie cognates?

Moreover, proto-Indo-European might even have retained protoforms for exotica such
as the monkey and elephant, at least according to Gramkrelidze and Ivanov (1995),
which, if we are to accept the evidence of such reasoning ex silentio, further problematizes
the European and Russian homeland theories and could even be used in support of the
case of the Indian homeland if we are to follow the same logic that has been levied
against it. Gamkrelidze and Ivanov (1995) reconstruct Indo-European animal words for
wolf, bear, leopard, lion, lynx, jackal, wild boar, deer, wild bull, hare, squirrel, monkey
and elephant. 17 Contrary to Thieme’s objection mentioned earlier, we find that, accord-

114 The Quest for the Origin of Vedic Culture

ing to Gamkrelidze and Ivanov, items unique to India actually might have cognates
elsewhere. They claim that dre monkey (Skr. kapi < * qhe/oph) has widely distributed
cognates (442). 18 Sanskrit ‘elephant’ also shares a cognate form with Latin ‘ivory’ (Skt.
ibha, Latin ebur < *yebh- or *Hebh-). Hittite-Luwian, and Greek, point to another protoform
( *lebh-onth- or * leHbho-), which suggests to Gamkrelidze and Ivanov that the two words
are related to a single Proto-Indo-European form for this animal. 19 Likewise, although
there are a variety of terms for lion, Dolgopolsky (1987, 10) considers the form *singh
as one of the few Proto-Indo-European animal terms that appear to be fairly reliable on
the basis of Indie (siriiha) and Armenian (inj ‘panther’).

Clearly, there are problems with some of the arguments oudined here; Gamkrelidze
and Ivanov’s reconstructions are by no means universally accepted—the terms for el¬
ephant and monkey may have been loans into later languages. Moreover, few Indo-
Europeanists still champion the beech or salmon evidence. But the point is that if beeches
or salmon or any other item can be promoted as proof of an European or Russian home¬
land, there is little to prevent disenchanted Indian scholars from coopting Gamkrelidze
and Ivanov elephants and monkeys in support of a South Asian one in order to dem¬
onstrate the maleability of this method. However, even allowing all of the arguments by
Dhar and others noted here, unless a few unambiguous inherited cognates among the
Indo-European languages for items unique to South Asian can be found, it is unlikely
that claims for a South Asian homeland will attract any serious attention from Indo-
Europeanists. Some cognates of tribal names from the Rgveda, at least, would be ex¬
pected to surface in the West if Indo-European tribes had emigrated there from India
(and if the Rgveda is as old as Indigenists would have it). Of course, as Gamkrelidze
and Ivanov’s reconstruction has shown, surprises are always possible (although not always
accepted), particularly when the data are approached with different perspectives; but,
with the exception of Elst and Dhar, linguistic paleontology remains another aspect of
this issue almost completely ignored by the Indigenous School.

Witzel, in the earlier remark about the elephant (hastin ‘possessing a hand’), articu¬
lates a further, often encountered observation regarding the names of some animals in
India: they are coined terms, newly formed from Sanskritic elements, as opposed to
terms formed from a primitive Indo-European verbal root. Masica (1991) elaborates:

Although spokesmen for the traditional Indian view try to fight back with selective mod¬
ern arguments, the philological evidence alone does not allow an Indian origin of the
Aryans. . . . the names of things peculiar to India . . . are for the most part either bor¬
rowed or coined (rather than “primitive”), cither of which may be taken as an indication
that the thing in question is new to the speakers of a language. (38)

Again, Elst takes (1996) objection to this: “Far from being an indication of more recent
and ‘artificial’ coinage, these descriptive nouns are the typical PIE procedure for creat¬
ing names for animal species” (380). He notes that Proto-Indo-European *bheros ‘brown’
has yielded the name bear; *kasnos ‘the gray one’, hare; *ekwos (which linguistis would
nowadays reconstruct as ‘fqekwo-) ‘the fast one’, horse; he argues that these are all crea¬
tures with accepted Proto-Indo-European pedigree, yet their nomenclatures consist of
‘coined’ rather than ‘primitive’ terms. Of course, as was pointed out in chapter five, one
has to see which words fit the appropriate Indo-European phonemic pattern, but, as we
will encounter repeatedly with Indigenous Aryan arguments, Elst simply reverses the

Linguistic Paleontology 115

logic of those supporting the Aryan invasions to conclude that “the argument from the
colourful descriptive terms in the Indo-Aryan languages will, if anything, rather plead
in favour of the IUT [Indian Urheimat Theory] than against it” (382). As for Dhar
(1930), he seems bewildered by such logic: “One fails to understand what has the ad¬
mission of Aryas into India got to do with the appellative name Hasti. Why could not
the Aryas be natives of India and at the same time give the elephant a name . . . ‘animal
with a hand’ . . . having been struck naturally by the animal’s unique and prominent
trunk?” (44).

Dhar has a point; even though -in suffixes are late derivations, Sanskrit does have an
old term for elephant, ibha, which it shared with a Latin cognate and might even have
been, at least according to Gamkrelidze and Ivanov, Proto-Indo-European: so the Indo-
Aryans had no need to invent a new term. It is quite likely that the word hasti is argu¬
ably a secondary, later, popular, folk term that gained currency. As Mallory (1975) re¬
marks, “Would we lay the blame to a non-Germanic substrate should Dobbin or Rover
replace ‘horse’ and ‘dog’?” (142). This comment is also relevant to the previous chapter
on substratum; language is never static, old terms get dropped, and new terms are coined
to replace them, but this need have nothing to do with immigration into a new, unfa¬
miliar landscape.

The Horse

Gamkrelidze and Ivanov’s book, (1995), although not without its critics, is the most
comprehensive recent work on linguistic paleontology. Actually, there is not much in
their reconstructed PIE environment that would compel Indigenous Aryanists to change
their views. Northwest South Asia contains many of the features that these scholars
have assigned to the homeland: it is certainly mountainous and forested, possesses moun¬
tain lakes and fast-rushing streams, can be characterized by cloudy skies with frequent
thunderstorms, is subject to heat and cold/ 0 knew herding and agriculture from the
seventh millennium b.c.e., contains most of the animals listed by Gamkrelidze and
Ivanov, produces honey, and certainly had a developed water transportation system by
the third millennium b.c.e.

The most pressing item from Gamkrelidze and Ivanov’s reconstructions that is likely
to be raised as an objection against an Indian homeland is the much later appearance
of the horse in the South Asian archaeological record as opposed to its much earlier use
in the steppes, where it was domesticated six thousand years ago (Anthony, Teiegin,
and Brown 1991, 94). Apart from one or two reports of early horse bones, which will
be discussed in chapter 9, the earliest evidence of horses in the Indian subcontinent is
generally dated to around the first half of the second millennium b.c.e. In the opinion
of many scholars, this paucity of horse bones in India indicates that the Indo-Aryans
entered this region well after dispersing from their original homeland. The horse evi¬
dence has long favored the Russian steppe homeland hypothesis and is die mainstay of
Gimbutas’s homeland theory. The horse has been the primary animal for which schol¬
ars have tried to account in the homeland quest, since it is culturally central to the vari¬
ous Indo-European traditions and was clearly known to the undivided Indo-Europeans.
Beekes (1995) finds the horse an “essential clue” providing “concrete evidence” from

116 The Quest for the Origin of Vedic Culture

the “facts” provided by linguistic paleontology that otherwise “don’t give us very much
to hold onto” (50). The horse is an essential part of the Indo-European world.

Accordingly, Mallory (1989, 163) immediately eliminates the Balkans and all other
areas where the horse was a late arrival from serious consideration as possible home¬
lands. Indian detractors of the Indigenous Aryan school, such as R. S. Sharma (1995)
and Shireen Ratnagar (1996b), also lean heavily on this late arrival of horse bones on
the subcontinent in support of their views. This lacuna in the Indian archaeological
record tends to haunt any attempt to argue for an Indian urheimat, and even (as will be
discussed in chapter 9), any efforts to correlate the Indus Valley Civilization with the
Vedic culture, which is a horse-using one. Since this animal has become almost synony¬
mous with the Vedic Aryans and, by extension, the whole Indo-Aryan migration de¬
bate, the horse evidence has to detain us at length, both here, in terms of linguistic
paleontology, and in the chapter on the Indus Valley, in terms of the archaeological
record.

When all is said and done, however, even the Proto-Indo-European status of this
animal is not without problems. There seems to be a recurring opinion among linguists,
going back at least to Fraser (1926), that considering *ekwos to have been a domesti¬
cated horse involves accepting some assumptions that can be called into question: “The
significance attached to the fact that the Indogermans were acquainted with the horse . . .
may have been exaggerated. We do not know the precise meaning of the Indogermanic
words in question; we do not know whether they mean the domesticated or the wild
animals.” For these types of reasons, “it is difficult to see how these names can be safely
used for determining the original home of the Indogermans” (266-267). D’iakonov
(1985) has reiterated this point more recently: “The Proto-Indo-European term for ‘horse’
shows only that horses were known (nobody doubts this); it does not mean that horses
were already domesticated” (11 3). Dolgopolsky (1990-93), noting the denotative vague¬
ness of the term, argues that in horse-breeding cultures there are words for ‘mare’ and
‘foal’. The fact that these terms cannot be reconstructed in Proto-Indo-European sug¬
gests to him that the referent of *eku>os must have been a wild horse (240-241); (how¬
ever, some linguists do reconstruct a term for mare, or, at least, that a word for ‘mare’
would have been expressed by a word for ‘horse’, coupled with an indicator of feminine
gender as in classical Greek. 21 Zimmer (1990a) points out that the inference that the
horse was known to the Indo-Europeans is primarily based on such poetic formulas as
‘swift horse’, ‘horses of the sun’, ‘characterized by good horses’, and so on. He feels
that “the formulas tell us nothing specific about the use of horses, but archaeology and
history supply the necessary information” (316). This observation is significant. Diebold
(1987) has elaborated on these points:

IE linguistics can agree on the reconstructed Proto-Indo-European etyma 'elcwos ‘horse’. . . .

But let us note [that) the animal terms tell us, in and of themselves, nothing about the
cultural uses of those animals or even whether they were domesticated; but only that Proto-
Indo-European speakers knew of some kind of horse . . . although not which equid. . . .
The fact that the equid *ekwos was the domesticated Equus caballus spp. Linnaeus . . .
cornels] not from etymology but rather from archaeology and paleontology. The most we
can do with these prehistoric etyma and their reconstructed proto-meanings, without ar¬
chaeological and paleontological evidence (which does indeed implicate domestication),
is to aver a Proto-Indo-European familiarity with these beasts. (53-54)

Linguistic Paleontology 117

There is an element of circularity with the horse evidence. Linguistics cannot tell us
whether Proto-Indo-European *ekwos known to the Proto-Indo-Europeans was the do¬
mesticated Equus caballus Linn, or whether it referred to some other species of wild
equid: the archaeology of the homeland does. But the archaeology of the homeland is
primarily located in the Kurgan area because that is where Equus caballus Linn was first
domesticated (an occurrence supposedly confirmed by linguistics)! Understandably, such
logic will hardly assist in convincing those already suspicious of the steppe homeland.
Since northwestern South Asia is the home of Equus hemonius khur, an equid sub¬
species called onager, I have even encountered the argument, using the logic outlined
earlier, that Proto-Indo-European *ekwos might just as well have originally referred to
a northwestern, South Asian hemonius khur, which was then transferred onto other
types of equids by outgoing Indo-European tribes leaving an Afghanistan/South Asian
homeland—although this is unlikely, since the word seems to have been generally ap¬
plied to denote the horse and not to donkeys or other equids (with the exception of Arme¬
nian where the cognate es does mean donkey). 22 Again, such possibilities are relevant not
as serious proposals suggesting that such an occurrence actually happened but as illustra¬
tions of how the assumptions involved in linguistic paleontology can be challenged and
reversed.

It is also important to note that, according to Dogra (1973-74), “only once do we
hear of actual horse riding in the Rgveda (V. 61 -62)” (54). McDonnell and Keith ([1912]
1967) note one or two other probable references to horse riding involving terms for
whips and reins (while remarking on some difference of opinion among scholars in
this regard), but they stress that there is no mention of riding horses in battle, which is
the image that has always been promoted by advocates of the classical Aryan invasion/
migration theory. Also of relevance is the fact that no words for typical riding equip¬
ment such as cheek pieces or bridles can be reconstructed for Proto-Indo-European, nor
is there a Proto-Indo-European etymon for horse riding (Zimmer 1990a, 321). Ivanov
(1999) notes that “if horseback riding really did began at the turn of the IV mil. B.c.
before the dispersal of Proto-Indo-European, it did not leave traces in the vocabulary of
the later dialects. . . . Thus it cannot be proven that this type of ancient , . . horseback
riding had originally been connected with Indo-Europeans” (233). 23

Coleman (1988, 450) notes that five different roots are attested for the animal in
various Indo-European languages. This suggests to him that either the protolexicon con¬
tained several words for horse depending on its function or that the animal was known
only in some areas of Proto-Indo-European speech, with the principal reconstructed
original word *eku>os being a dialectal one, and the other words innovatory after the
dispersal. Along similar lines, Lehmann (1993, 272) argues that the fact that there is
only one solidly reconstructable word (*ekwos) for an item that was of such centrality to
a culture further underscores the lateness of the borrowing into Proto-Indo-European.
He notes that modern terms of equivalent centrality, such as automobile, are known by
a myriad of terms. The generic term is initially adopted and then various languages
innovate their own names for the item. But only one term is reconstructable for horse
in Proto-Indo-European underscoring the fact that it had not been in the protolexicon
for very long before the dialects dispersed.

Alternatively, Lehmann also argues that *ekivos could have been a later loanword
that circulated throughout the various dialectics after their dispersal, perhaps being

118 The Quest for the Origin of Vedic Culture

phonemically restructured in some areas. This generic loan was maintained along with
other terms that arose in time in individual languages. Lehmann (1993, 271) refers to
the many phonological problems, as well as desperate solutions, incurred by linguists
in attempting to reconstruct the protoform for this term and suggests that the word may
have been a borrowing that was adjusted to fit the phonemic pattern typical of indi-
vidual dialects or dialect groups. 24 He illustrates this possibility by means of the example
of batata ‘potato’ that was introduced into Europe in the sixteenth century. This word
was restructured as patata, pataka, patalo, tapin, katin, patal, and so fordt, simply in the
Romance languages alone. He notes that even these cognates within Romance show far
less diversity than the variants for equus in Indo-European. This is a relevant observa¬
tion. In the earlier phases of the expanding linguistic continuum of the Indo-European
languages out of the homeland, wherever that might have been, the dialectal differentia¬
tion within the continuum (which will be discussed in chapter 8) would not have been
as pronounced as in later periods. So the horse or wheel, for example, could have been
new items that were encountered at some point on the continuum, which were then
shared, along with their names, with the other Indo-European speakers elsewhere on
the continuum. These terms, although loans, would appear to be inherited since the
dialectal differentiation between the languages that adopted the terms might not be
sufficient to detect them as loans. Lehmann’s comments could support such a possibil¬
ity, especially since he holds that the dialectal differentiation is indeed sufficient to iden¬
tify this word as a loan, which might mean that the loan circulated at a later time when
the dialects had differentiated more.

If these linguists are corrent that the word for horse could have circulated after the
dispersal of the Indo-Europeans, and then been restructured according to individual
dialects, then stating that the Indo-Europeans knew the horse before their dispersal and
therefore must have inhabited an area wherein the horse is native (and eliminating other
areas where the evidence for the horse is a later phenomenon) becomes less convincing.
Indeed, the corollary of all these arguments suggesting that *ekwos is either a late Proto-
Indo-European word or a loanword that circulated after the dispersal of the Indo-Euro¬
pean languages is that the homeland could not have been in the steppes where the horse
is native. Had the homeland been there, the Proto-Indo-Europeans would have always
been around horses and would have had an ancient word for the animal in their lexi¬
con, and not a more recent or a restructured one.

However, not all linguists would agree that the word is either late or borrowed. If we
accept that the word is inherited from the proto-period, and accept that the Indo-Europeans
were an undivided entity until somewhere between 4500 and 2500 b.c.e., as most scholars
would hold, then we have anything from about a one to three millennia gap between
when the horse was known to the Indo-Europeans and when it is unambiguously evi¬
denced in the South Asian archaeological record. This is irrespective of whether the
horse was domesticated or wild. How can a South Asian homeland account for this?

Allowing that *ekwos does refer to a domesticated caballus Linn, the most convinc¬
ing argument used by the Indigenous Aryan school to account for its absence in the
subcontinent is that horse domestication may well have occurred in the steppes, since
this is the natural habitat of the animal, but it is an unwarranted assumption to then
conclude that the Indo-European homeland also must have been in the same area. As
D’iakonov remarks: “The Proto-Indo-European term for ‘horse’ shows only that horses

Linguistic Paleontology 119

were known” (113). Indeed Ivanov (1999), who has undertaken by far the most com¬
prehensive study of the cognate terms for horse in Indo-European as well as the adja¬
cent languages of Northern Caucasian and Hurrian, points out that “the Indo-Euro¬
pean homeland need not be identical to the area of horse domestication, but should be
connected to it. The ways in which names and technical knowledge . . . spread should
be explored” (1971). Thus Talageri (1993), argues that “the horse could have been very
well known to the proto-Indo-Europeans in their original homeland before their dis¬
persal from it (which is really the only thing indicated by the facts), without the horse
necessarily being a native of that homeland, or they themselves being its domesticators”
(158). The horse, according to this line of argument, was an import into India—a highly
prized, elite item. The paucity of horse bones in the early archaeological record is due
precisely to the fact that the animal, although highly valued, was a rare commodity used
in elite priestly or military circles. According to the horse specialist Bokonyi (1997):

It is well known that wild horses did not exist in India in post-Pleistocene times, in the
time of horse domestication. Horse domestication could therefore not be carried out there,
and horses reached the Indian subcontinent in an already domesticated form coming
from the Inner Asiatic horse domestication centres via the Transcaspian steppes, North¬
east Iran, South Afghanistan and North Pakistan. The northwestern part of this route is
already more or less known; the Afghan and Pakistani part has to be checked in the
future. (300)

In fact, the horse has always been highly valued in India. From the Vedic, through
the Epic, and up to the Sultanate period, it has always been an elite item, and it has
always been an import. According to Trautmann (1982), “the supply of horses . . . has
been a preoccupation of the rulers of India, from, nearly, one end of its recorded his¬
tory to the other. ... It has yet to be determined why exactly India has never been self-
sufficient in horses. Climate? A relative scarcity of pasture?” For our purposes, the fact
remains that “whatever die reason, the stock has always had to be replenished by im¬
ports, and the imports came from westward in the ancient period. ... It is a structure
of its history, then, that India has always been dependent upon western and central
Asia for horses” (261).

Elst (1996) ruminates on what a prehistoric scenario might look like from an Indig¬
enous Aryan point of view:

The first wave of Indo-European emigrants . . . may have reached the Caspian and Black
Sea coasts and domesticated the horse there, or learnt from the natives how to domesti¬
cate the horse. They communicated the new knowledge along with a few specimens of
the animal to their homeland . . . along with the appropriate terminology, so that it became
part of the cultural scene depicted in Vedic literature. Meanwhile, the Indo-European
pioneers on the Black Sea made good use of the horse to speed up their expansion into
Europe. (40)

The logic here is that the horse is highly prized in all the literary records throughout
Indian history, but it has never been indigenous (although foreign breeds have been
imported and bred on the subcontinent with varying degrees of success in the North¬
west-later Vedic texts speak about the fine horses of Kandahar and other places). 25 That
it was central to the Vedic texts, despite not being indigenous, is therefore no indica¬
tion of the indigenousness of the Indo-Aryans themselves—it was the horse that was

120 The Quest for the Origin of Vedic Culture

imported, not the Indo-Aryans. Elst is extending this same logic to argue that all the
Indo-Europeans could have been situated in India, where the horse was a highly prized
but imported and rare luxury item in the collective Indo-European consciousness. The
animal was encountered in the steppe area by the northwesternmost border of the Indo-
European language continuum that was expanding out of India. The creature, and its
name, were then relayed back to other areas of the language continuum, including In¬
dia. Lehmann observed drat the term could have been restructured according to the
various dialectics that were germinating in this continuum, making the term appear
inherited rather then a loan. Since the horse was such a useful creature, but also a rare
one, it became a much prized item in the Vedic sense.

While all this may be possible from a linguistic point of view, from an archaeologi¬
cal perspective the burden of proof will still remain with anyone who proposes that a
particular animal or item existed in the proto-historic period. As will be discussed in
the chapter on the archaeology of the Indus Valley, the horse remains a problem for the
Indigenous position even if, as Renfrew (1999) and others hold, “the significance of the
horse for the understanding of the distribution of early Indo-European has been much
exaggerated” (281).

Other scholars have tried to compensate for the lack of horse bones in India by coun¬
tering that the Russian steppes also lack faunal remains that are clearly Indo-European.
Such negative evidence was used by patriotic European-homeland promoters in the nine¬
teenth century to reject an Asian steppe homeland because of the absence there of another
important IE creature with very well attested cognates—the bee (and its honey). Dhar
rejected the steppes of southern Russia because they were unsuitable for agriculture.
Gamkrelidze and Ivanov have used the presence of exotic items such as the monkey and
elephants in Indo-European as evidence also opposing a Russian homeland. Most recently,
as discussed in chapter 1, Krell (1998) has produced lists to argue that the Kurgan area is
significandy incompatible with the evidence of linguistic paleontology. Using such nega¬
tive evidence, by the same logic used to eliminate India as a candidate, ultimately any
potential homeland can be disqualified due to lacking some fundamental Proto-Indo-Eu¬
ropean item or another. In addition, it has long been pointed out that this use of negative
evidence is suspect, since we have many examples of other flora and fauna that must have
been known to the earliest Indo-European-speaking communities but show minimal or
no evidence of cognates across Indo-European stocks (Mallory 1997).

Criticisms of the Method

Linguists have long been aware of the speculative nature of linguistic paleontology: “All
prehistoric reconstruction is of course purely hypothetical, that is, based on conjectural
assumptions. Stricdy speaking any conjectural assumption is a guess. ... A prehistorian
depends on . . . his imagination . . . trained by experience” (Thieme 1964, 585); “The
apparent existence of a common term in the language, which is attained through recon¬
struction on the basis of the attestations in the daughter languages, does not prove drat
the item it denotes actually existed in the relevant original society” (Polome 1992, 370). 26
Other scholars have been much more radically dismissive of the whole premise of recon¬
structing a hypothetical language and culture on the basis of cognate words present in

Linguistic Paleontology 121

textual or spoken languages existing thousands ofyears later. The linguist J. Fraser (1926),
for example, presented a well-known (but faulty) caricature of the whole enterprise by re¬
constructing a proto-Romance scenario from the paleolinguistic evidence of the historic
Romance languages: “By th[is] same method of investigation we shall discover that the
Romans had emperors, and a republic; drat drey had priests, called by a name represented
by the French pretre, and bishops; that drey drank beer, probably, but certainly coffee, and
that they smoked tobacco” (269). Unfortunately, biere, tabac, and cafe are all loanwords,
minimizing the persuasiveness of Fraser’s caricature (which was more accurately a recon¬
struction of the Vulgar Latin of much later times), but linguists disillusioned with this
whole method nonetheless supported dre spirit of his critique: 27

Now the more sophisticated among us could easily object here that it would take a great
deal of naivete on the part of linguistic palaeontologists to propound such views, . . . yet
such naivete seems to enjoy the status of high acumen, as anyone can see who reads some
of the numerous volumes that deal widt the “Indo-Europeans,” their lives and their mo¬
res. But if the authorship of such works is not astonishing enough, the uncritical and
admiring credulity bestowed upon them by a vast number of scholars certainly is. (Pulgram
1958, 147) 28

Latinists in particular, like Lazzeroni, supported Pulgram’s punto fundamentals and rejected
the capability of linguistics to ever be able to determine where a protolanguage was spoken,
even if it could reconstruct portions of what was spoken.

Fraser (1926) also pointed out that words in language A, associated with particular
geographic locations, travel freely and are borrowed by speakers of language B outside
those locations. Such words, if found by the linguistic paleontologist, could be erroneously-
interpreted as indicating that the speakers of language B originated in the territory of lan¬
guage A. These criticisms are well worth quoting, bearing the horse evidence in mind:

The English language has laid under contribution almost every language on the face of
the earth. We speak freely of the fauna and flora of other countries, not merely [of]
England. . . . Names like ‘lion’, ‘tiger’, ‘wine’, ‘cotton’ . . . and hundreds of other things
which are not indigenous in England but are perfectly familiar to every speaker of the
English language all over the world. ... 1 do not see how scholars placed in the same
relation to English as we are to Indo-Germanic could tell that the Englishman knew
cotton, wine, and the like only through literature or as articles of commerce, and not
because he lived in a region which produced them all. That the palaeontologist of the
future . . . should describe the Englishman as tending his vines in the neighborhood of
tiger-infested jungles, would not, perhaps, be very astonishing. (272)

There are, of course, ways that a linguist can, in some cases at least, determine whether
a word is a later loan or an item inherited from the protolanguage. In the latter case, the
phonemic or morphological properties of the word and its root structure would be ex¬
pected to be consistent with the rest of the language in which it occurs. As noted in
chapter 5, the basic axiom is that if a word is inherited from the protolanguage, it must
show all of the sound shifts that its phonemes should have undergone over the relevant
period of time (i.e., that other words in the language with similar phonemes and in
similar linguistic contexts would have undergone) according to the rules of evolutionary
sound and change. If any phonemes in the word fail to demonstrate a required sound
shift, or if the word shows a non-I.E. root structure, then it is a good candidate for

122 The Quest for the Origin of Vedic Culture

consideration as a loanword. Nonetheless, Fraser’s objections are still echoed by schol¬
ars such as Renfrew (1987, 1988) in his scathing (and criticized in turn) critique of
linguistic paleontolgy, to the approval of Coleman (1988), and should be kept in mind
when building up a theory overly dependent on cognate words. As Lehmann (1993)
and others have argued, Proto-Indo-European *ekwos could have been a new innova¬
tion or loanword that, along with knowledge of horse use, spread throughout the dis¬
persed Indo-European- language-speaking area after the breakup of Proto-Indo-European
(although not later than the stage of Old Indo-Iranian). 29

Such inherent imprecision of linguistic paleontology has provoked consistent criti¬
cisms of the method by linguists. Keith (1933) had long pointed out that:

The determination of the Indo-Euopean civilization is precisely the point which affords
least hope of any satisfactory result. It rests on linguistic evidence pure and simple, and
it is open to the gravest doubt whether such evidence is capable of giving the results which
are claimed for it by those who seek to determine die Indo-European home. ... It should
suffice to remember that on the basis usually adopted we would have to conclude that the
Indo-Europeans knew snow and feet, but were ignorant of rain and hands. The difficulty,
of course, is in theory recognized by all who deal with the issue; the trouble is that in
practice they tend more or less completely to ignore it, and to create for us a picture of the
Indo-Europeans which is probably a mere delusive shadow of the actual civilization of
the people. Yet it should be a warning when we find that linguistically we may assert that
the Indo-Europeans knew butter but were unacquainted with milk. (189-190)

Pulgram (1958) was even more dismissive:

It is an elementary mistake to equate common Indo-European words with Proto-Indo-
European words and to base thereon conclusions concerning the Proto-European Urvolk
or Urheimat. Yet this is precisely what has often been done. . . . impassioned linguistic
palaeontologists have gone even further. From the existence of certain items of vocabu¬
lary in all or a majority of the extant Indo-European languages, and blandly ignoring all
the pitfalls just noted, they even fabricated conclusions concerning the social organization,
the religion, the mores, the race of the Proto-Indo-European. (Pulgram 1958, 145-146)

More recently, Coleman (1988) while noting that some progress had been made, and
more could be expected from the method, nonetheless confers that “the arbitrary and
unrigorous methods that have characterized much of this linguistic paleontology cer¬
tainly deserve Renfrew’s scepticism. . . . Most of the lexemes that can be confidendy
assigned on the basis of widespread attestation ... do not tell us much” (450). McNairn,
commenting on Kossina’s and Gimbutas’s employment of the method, remarks that
“the clues afforded by linguistic paleontology were either so general that they accommo¬
dated both centres without much difficulty, or they were so hypothetical that they could
be easily ignored if unsuitable” (quoted in Anthony 1995b, 96).

Most recently, Krell (1998) argued that “the old, pliable crutch of linguistic paleon¬
tology should certainly be abandoned, at least until the theoretical uses and limitations
of the Proto-Indo-European lexicon have been more precisely defined” (280). She points
out that the reconstructed lexicon does not provide a linguistic picture of a group of
Proto-Indo-European speakers at one point in time, or even in one location in space; it
may well represent a linguistic continuum of several millennia into which different lexi¬
cal items were introduced at different stages. Most important, she reiterates the com-

Linguistic Paleontology 123

mon objection that it is virtually impossible to identify the exact or even approximate
referent of a reconstructed lexical item: “It is imperative, in working with the problem
of Indo-European origins, that the contents of the PIE lexicon not be treated too literally.
Historical linguistics has shown numerous examples of how dramatically the meaning
of a given word can shift in the course of a few centuries, let alone several millennia”
(279). She concludes that “the use of so-called ‘linguistic paleontology’ . . . has always
been a popular method in the construction of Proto-Indo-European urheimat theories.
It rests entirely on the supposition that the meaning of a proto-form can be reconstructed
beyond a reasonable doubt, a supposition which I argue is false” (279).

Conclusion

Despite such inherent problems, theories about the Indo-European homeland are still
sometimes predicated on linguistic paleontology for their geographic identification. 30 Thus
Mallory (1989) first uses the method to delineate a broad area and then concludes: “We
have pushed the linguistic evidence about as far as we may; now it is the turn of the ar¬
chaeologists” (165). Clearly, however, anyone disenchanted by the initial linguistic method
is not likely to give much credence to the secondary auxiliary archaeological evidence that
might be called in for support. And archaeology by itself, as Talageri (1993) notes, echo¬
ing objections outlined in chapter 1 that have been repeatedly made, tells us nothing about
the language spoken by the members of a material culture. 31 One material culture does
not indicate one linguistic entity, nor does the spread of a particular material culture nec¬
essarily equate the spread of a particular language group, any more than the spread of a
language group corresponds to the distribution of a specific material culture (135).

One has only to glance through any of the various homeland hypotheses to see how
the same linguistic evidence is utilized very differently by different scholars. Depending
on one’s own perception of things, one will find alternative theories far too compli¬
cated. Any claimant for the homeland has to engage in special pleading, or at least feels
that other contestants are more extreme pleaders, which underscores the limitations of
the method of linguistic paleontology. The judgment on Occam’s razor is very likely to
be perceived quite differently in India than in the West.

Ultimately, the dramatically different scenarios still arrived at by different scholars using
linguistic paleontology are, in themselves, sufficient proof of its unreliability, if not inade¬
quacy, at least in its present state. If Gimbutas is satisfied that linguistic paleontology can
support the reconstructed Proto-Indo-European par excellence as an aggressive, mounted,
nomad warrior where Renfrew (albeit dismissive of the whole method) believes it cannot
exclude his gende, sedentary agriculturist, or if the method can be used to promote the
environment of the proto-urvollc as the harsh, cold, and austere northern one of the steppes
but yet be adopted by Gamkrelidze and Ivanov to suggest a warmer and more exotic southern
one with tropical animals, then obviously something is inadequate widr the present state
of the method. If the method is so problematic or limited in reliability, and treated sceptically
or rejected even by most present-day Western linguists, one is forced to question how it
can be used as conclusive or even persuasive evidence to compel disenchanted Indian
scholars to believe in die theory of Aryan invasions or migrations into the Indian subcon¬
tinent predicated on this type of data.

Linguistic Evidence from outside of India

Potentially devastating evidence against the case of the Indigenous Aryanisrs is the exis¬
tence of loanwords between Indo-European, Indolranian, and Indo-Aryan, and non-
Indo-European language families. These suggest the geographic proximity of the Indo-
Europeans and/or the Indo-Aryans with other language groups far from the horizens of
the Vedic Indo-Aryans. Loanwords are often taken to be an essential ingredient in geo¬
graphically tracing the prehistories of language families, since, if they occur in sufficient
numbers, they suggest that the families in question were once situated adjacent to each
other. Accordingly, if Proto-Indo-European contains loanwords from language families
far from South Asia, then these loans provide compelling evidence that Proto-Indo-
European could not have originated in South Asia. The same method applies to Indo¬
lranian and Indo-Aryan. Loans between these and other languages provide relevant data
for attempting to chart respective points of origin, migrations and trajectories of these
languages.

Semitic Loans in Indo-European: Nichols’s Model

The detection of such loanwords is the primary' method used by a number of linguists
who locate the homeland in the Near East. Gamkrelidze and Ivanov, for example, to a
great extent base their homeland thesis on the number of loanwords, particularly from
Semitic and South Caucasian (Kartvelian), that they trace to Proto-Indo-European. 1 This
evidence suggests to them that the Proto-Indo-Europeans must have been situated adjacent
to the Semitic and Caucasian language families, somewhere in the vicinity of Armenia.
Etymologizing rarely produces consensus, as we have already seen, and most of Gamkrelidze

124

Linguistic Evidence from outside of India 125

and Ivanov’s etymologies have been challenged in one way or another by D’iakonov, an
advocate of a Balkan homeland, but supported by Dolgopolsky, who also accepts a Near
East homeland, albeit more in central Anatolia (and with different dialectal maneuverings
therefrom). Shevoroshkin likewise supported a homeland in the eastern part of Asia Minor
but differed from Gamkrelidze and Ivanov about whether there were significant loans
between the north and south Caucasian languages and Proto-Indo-European and whether
these language groups were therefore immediately adjacent to each other. 2

With regards to such Semitic and Caucasian loanwords in Proto-Indo-European, Nichols,
who translated Gamkrelidze and Ivanov’s magnum opus from Russian into English, of¬
fers a theory that is particularly relevant to our line of inquiry. Her methodology involves
tracing the linguistic history of Semitic cultural loanwords from urban and cosmopolitan
Mesopotamia into the surrounding areas (which might be considered to have had “lower”
culture and therefore to have been more prone to borrow). Loanwords often emanate out
from a central area—especially if this area is perceived as a “higher” or more prestigious
urban culture—and are borrowed by adjacent languages, which can rephonemicize the words
according to their own phonetic system. Nichols notes that these loanwords may then be
passed on in turn to the other neighbors of these languages (which are thereby not imme¬
diately adjacent to Mesopotamia but twice removed). These further rephonemicize the words
according to their own sound systems. An expert linguist could, at least theoretically, trace
those words’ history, which could be indicative of the relative geographic situation of all
these languages, and in particular of Proto-Indo-European. The assumptions here (which
can obviously be called into question, since loans can travel vast distances through trade
or other means) are that significant loanwords between language groups indicate geographic
proximity of these groups, and lack of loanwords indicates that the languages in question
were not immediately adjacent to each other. ?

In terms of linguistic geography, Nichols translates these findings into the following
conclusions: Proto-Indo-European could not have been situated between Mesopotamia

Northeast Caucasian
(Abkhaz-Circassian)

Northwest Caucasian

(Nakh-Daghestanian) , . _

Indo-European

Nichols’s schematic rendition of Indo-European in relation to other languages (after
Nichols, 1997).

126 The Quest for the Origin of Vedic Culture

and the Black Sea coast, since culturally laden vocabulary emanating from Mesopotamia
did not surface in the language spoken in this coastal area, West Caucasian (Abkhaz-
Circassian), via Proto-Indo-European. 4 Proto-Indo-European was not situated between
Mesopotamia and the eastern Caucasian foothills and Caspian coastal plain for similar
reasons: the language spoken there, East Caucasian (Nakh-Daghestanian), also shows
no sign of Proto-Indo-European loanwords. As an aside, and irrespective of any pos¬
sible Semitic loans from Mesopotamia, we can note that if Proto-Indo-European did not
impart any of its own native words either into the languages spoken on the Black Sea
Coast or into those spoken in the eastern Caucasian foothills, then the claim of geo¬
graphic distance between Proto-Indo-European and these languages is reinforced, at least
from the perspective of this method. These data are further obstacles for those propos¬
ing a steppe homeland. 5

Be that as it may, at this point Nichols deduces that Proto-Indo-European did not lie
to the northwest or northeast of Mesopotamia. It did lie in a direct trajectory of Semitic
and Sumerian loanwords but not one immediately adjacent to Mesopotamia, since, accord¬
ing to Nichols, the Semitic loans in the protolanguage show signs not of direct borrow¬
ing but of filtration through an intermediary. 6 Nichols accordingly situates Proto-Indo-
European still farther to the northeast, in Bactria-Sogdiana, since that is where it could
spread across the steppe. 7 She reinforces dris with an innovative model of language spread
and dialect geography which will be discussed in chapter 8. This location, needless to
say, would be welcomed by the Indigenous Aryanists since it overlaps the area under
consideration here, namely, an Afghanistan/Pakistan/Northwest Indian locus of origin. 8
I will return to Nichols later.

Finno-Ugric Loans

Of more pressing significance to the Indo-Iranian languages are the loanwords that have
been transmitted from them into the Finno-Ugric language family, which was probably
spread throughout northern Europe and northwestern Asia in the prehistoric period. 9
Finno-Ugric contains numerous loanwords that, depending on the linguist, have been
identified as either Indo-Iranian, Iranian, or Indo-Aryan, indicating that these languages
must have been adjacent to each other in prehistoric times. Since there is absolutely no
evidence suggesting the presence of Finno-Ugric speakers near the Indian subcontinent,
it is reasonable to conclude that Indo-Iranian speakers must have been present in north¬
west central Asia. How, then, could they have been indigenous to India or, even, the far
Northwest of the subcontinent and Afghanistan? The conventional explanation for this
is that the Indo-Iranians, after leaving their original homeland wherever that might have
been, sojourned in areas adjacent to the Finno-LJgric speakers before proceeding on to
their historic destinations in Iran and India.

S. S. Misra (1992) has offered a rather different explanation. Misra draws attention
to one rather significant feature regarding these loanwords, which he believes is deci¬
sive in determining the direction of language flow corresponding to Indo-Aryan move¬
ments: the loans are from Indo-Iranian into Finno-Ugric. There are no loans from Finno-
Ugric into Indo-Iranian. This is a crucial point. Misra argues that had the Indo-Iranians
been neighbors with the Finno-Ugrians in the regions to the north of the Caspian Sea

Linguistic Evidence from outside of India 127

The Finno-Ugric language family.

for so many centuries, then both languages would have borrowed from each other. If
the Indo-Iranians, as per the standard view of things, had, then, journeyed on toward
their historic destinations in the East, they should have brought some Finno-Ugric loans
with them in their lexicons, at least a few of which should reasonably be expected to
have surfaced in the earliest textual sources of India and Iran. But, as Burrow noted
some time ago, it is usually quite clear that these words have been borrowed by Finno-
Ugric from Indo-Iranian and not vice versa (1973b, 26). 10 There do not seem to be any
Finno-Ugric loans evidenced in the Veda or the Avesta. This, for Misra, indicates that
the Indo-Iranians never went from the area neighboring the Finno-Ugrics down to Iran
and India; they went from India to the Caspian Sea area, where they encountered Finno-
Ugrians. The Finno-Ugrians, in this version of events, could therefore freely borrow
from the Indo-Iranians, but since those emigrating Indo-Iranians never returned to Iran
and India (at least in large enough numbers to affect the lexicon back home), no Finno-
Ugric loans ever surfaced in the Indo-Iranian literary sources. This version of events
accounts for the one-way borrowing. The argument is ingenious. However, as with every¬
thing else, counter arguments can be brought forth, such as the power dynamics of socio¬
linguistics (whereby a lower status group may borrow terms from a higher one without
the latter, in turn, borrowing terms from it). Moreover, Redei (1983) finds it “possible”
that the Uralic languages did not just borrow from the Aryan ones but also loaned them
words as well (15). Joki (1973), while criticizing the work of earlier linguists who had
attempted to find Uralic loans on the Indo-European side, nonetheless states that “in
his view, it is not impossible” that a dozen or so such cases can be argued in the Aryan
languages (373).

Since these loans are specifically from Indo-Iranian (and not Proto-Indo-European)
Dolgopolsky (1990-93) has employed a parallel logic (in support of his Balkan home-

128 The Quest for the Origin of Vedic Culture

land) to insist that Proto-Indo-European could never have been spoken in the steppes
north of the Caspian Sea:

What really matters is the fact that there are no proto-Indo-European loanwords in Uralic
(or Finno-Ugric) and no Uralic or Finno-Ugric loans in Proto-Indo-European. It strongly
suggests that there was no territorial vicinity between Proto-Indo-European and proto-Uralic
or proto-Finno-Ugric, that is that Proto-Indo-European was not spoken in or near the Volga
or Ural region, including the steppes to the North of the Caspian Sea (Gimbutas’s “Indo-
European homeland”). (242-243)

This, too, is a significant observation, particularly in view of the fact that Nichols has
pointed out that there were no Proto-Indo-European loans in the other languages bor¬
dering the steppes on the Black Sea coast and eastern Caucasian foothills. 11 While most
linguists seem to agree that the loans are Indo-Iranian and not Proto-Indo-European,
there is disagreement over whether they are specifically Indo-Iranian, Iranian, or Indo-
Aryan. 12 Misra (1992), in addition to reversing the direction of language flow, is of the
opinion that most of the words can be accounted for as Old Indo-Aryan forms and not
Iranian. Shevoroskin also considers them to be Indo-Aryan (and even Middle Indo-
Aryan). Most recently, Lubotsky (forthcoming) concurs that the oldest layer of borrow¬
ings are often of Sankrit and not of Iranian (6). D’iakonov (1985) and Dolgopolsky
(1993) consider them Indo-Iranian. Gamkrelidze and Ivanov (1983b) in contrast, are
quite specific that they should be interpreted as early Iranian and not as Indo-Iranian,
“or even less as Old Indie” (67). Joki (1973, 364-365) also considers them to be mostly
Iranian or Middle Iranian. As Mallory (1997) quips: “Will the ‘real’ linguist please
stand up? It should be obvious that linguists have as much difficulty in establishing the
chronological relationships between loanwords as any other ‘historical science’” (98).

The identification of the words as Indo-Aryan is crucial to Misra’s attempt to date
the Rgveda, His line of reasoning is predicated on the work of the Hungarian linguist
Harmatta, who analyzed these loanwords into eleven consecutive chronological periods
based on different stages of phonemic development. In Harmatta’s schema, the earliest
loans are Indo-Iranian, and later loans contain a variety of Iranian forms (recognizable
as Proto-Iranian, Old Iranian, and Middle Iranian), 11 stretching over a very long period
from the first half of the fifth millennium b.c.e . 14 to the Hun invasion of Europe in the
fourth century a.d.

In Misra’s analysis (1992), only the last (and chronologically latest) of the eleven
stages of Harmatta’s list contains forms clearly Iranian; all the earlier stages contain
Indo-Aryan forms. 15 Since Misra considers Old Iranian to have had a similar linguistic
and temporal relationship with Old Indo-Aryan as Middle Indo-Aryan had—namely,
that it was later—he concludes that the earliest loans were from the period before the
Iranians had split from the Indo-Aryans. Since Misra clearly considers proto-Indo-Iranian
to be very similar to Old Indo-Aryan, and since Harmatta had speculated that the oldest
Indo-Iranian loanwords occurred in the fifth millennium b.c.e,, Misra concludes that,
if this date is correct, Indo-Aryan must also be assigned to this period and therefore be
very much older than the commonly assigned date of 1200 b.c.e. One must note at this
point, that the value of Misra’s work is dependent on the value of that of Harmatta’s.
In any event, the dating of the Veda will be discussed in chapter 12 and is peripheral
to Misra’s essential point; for the purposes of the present discussion, Misra’s main ar-

Linguistic Evidence from outside of India 129

gument is that since the loanwords are only from Indo-Iranian into Finno-Ugric and
not vice versa, is just as likely, or even more likely, that the Indo-Iranians came from
South Asia or Afghanistan to the Caspian Sea area and not vice versa.

It must, be noted however, that a possible objection to both Nichols’s and Misra’s
east-to-west direction of language flow has been raised by some linguists (Shevoroskin
1987; Dolgopolsky 1989). These scholars hold that “the linguistic evidence, indeed, fully
corroborates the eastward direction of the migration of Indo-Iranians," which they argue are
evidenced by the chronological spread of loanwords from west to east (Shevoroskin 1987,
229; italics in original). The argument is that loanwords in the East Caucasian language
are the oldest and bear witness to Old Indo-Iranian, while the Finno-Ugrian family of
the lower Volga Valley has somewhat later loanwords from Middle Indo-Iranian. Accord¬
ing to Shevoroskin, loans from the Late Indo-Iranian linguistic period can be found in
languages emanating further eastward across the steppe. 16 These findings fit comfort¬
ably with their proposed Near Eastern homeland: “The historical implications of this
linguistic evidence are obvious: the proto-Aryans appear to move eastwards across the
Ponto-Caspian steppes, from the region north of the Caucasus, where they were in contact
with proto-Nakh-Daghestanian, into the Lower Volga area, where drey contacted proto-
Finno-Ugrians” (Dolgopolsky 1989, 29).

As we have seen with the Finno-Ugric borrowings, however, there is unlikely to be
agreement among linguists regarding the exact linguistic (and therefore chronological)
identification of such loans as Indo-Iranian, Indo-Aryan, or Iranian—what to speak of
more subtle distinctions such as whether they are Old, Middle, or Late Indo-Iranian.
Clearly, tracing chronological trajectories of loanwords from east to west or west to east
will entail levels of linguistic sophistication and an adequate degree of agreement among
linguists that is something we have yet to look forward to. And such research would
have to account for all the Indo-Iranian loanwords in the Caucasian and Ugric lan¬
guages, not just a select few. 17 Moreover, even if it could be determined that Nakh-
Dagestanian (East Caucasian) has older loanwords than Finno-Ugric from an unified
Indo-Iranian period, these data are not incompatible with a position similar to Nichols’s
model (1998) which will be considered in the next chapter. In her model, Older loanwords
could have belonged to an earlier wave of Indo-Iranian from an eastern locus of origin
which impacted Finno-Ugric, and this could have been followed by a later wave of Middle
Indo-Iranian, which influenced the more southerly Caucasian area. Ultimately, there is
little in the history of loanwords that can eliminate a variety of historical possibilities.

Other Traces of Indo-Aryan

The Finno-Ugrics were not the only other language family to borrow from the Indo-
Iranians, who seem to have left linguistic traces across large areas of Asia over a wide
time span. Where Nichols seems to have found Proto-Indo-European loans, Harmatta
finds evidence of a number of proto-Iranian or proto-Indo-Aryan loans in the Caucasian
and Ketic languages. The Iranians seemed to have gone as far as the borders of China
and Korea, since here, too, he finds Iranian loanwords; again Misra considers them to
be Indo-Aryan (on the grounds that they show s in places where Iranian shows h). Russian
linguists have also pointed out Indo-Iranian, or possibly Indo-Aryan, tribal names and

130 The Quest for the Origin of Vedic Culture

hydronyms in the Kuban region north of die Black Sea, as well as around the Caspian
Sea. 18 Telegin (1990) draws attention to thirty Iranian hydronyms in the Dnieper basin.
There are also Indo-Aryan and Iranian terms in European hydronomy (Schmid 1987,
331). The Russian linguist Karamshoyev (1981) finds Iranian words (to which Misra
agrees) in the Pamir languages of Afghanistan—the much favored homeland of the nine¬
teenth century.

Burrow (1973a) found evidence of Indo-Aryans in western Iran, soudi of the Caspian
Sea. The Avesta, which is geographically centered in eastern Iran, makes mention of the
Mazanian daevas, who are worshipers of the Indo-Aryan gods. According to Burrow, Mazana
is known in Iranian sources as the territory between die southern shore of the Caspian
Sea and the Alburz Mountains. These daeva worshipers, although still in existence at the
time of the inscriptions of Xerxes, were condemned by the reformer Zarathustra, perse¬
cuted, and eventually subsumed by the Iranians from the East. Parpola (1988, 127) notes
that in Latin sources, Strabo (11.9.2) refers to a people called Parnoi who belonged to the
Da(h)as, who were said to have lived in Margiana, from where they founded the Arsacid
empire of Parthia. Actually, in this instance, we have a clear indication of a movement
from east to west, since the Dahae came to live on the east coast of the Caspian Sea, north
of the Hyrcania (Gurgun), where there is still a district called Dahistan. Parpola notes that
the Parnoi corresponds to the Sanskrit term Pani (if this is accepted as a Prakrit develop¬
ment <*Prni). The Dahas/Da/iae correspond to the Dasa and the Panis to another tribe
that fought the Aryan Divodasa on the banks of the Sarasvati.

When the unambiguous Indo-Aryan names found in documents in the Middle East
(to be discussed later), are also taken into account, the picture that emerges through lin¬
guistic and philological sources is of Indo-lranian, Indo-Aryan or Iranian tribes spread
out over an extended period and across a vast area stretching from India to the Middle
East and across to the Great Hungarian Plain, up to the north of the Caspian Sea, and
over as far as the Ordos in northern China (Harmatta 1992, 359). These have invariably
been taken by Western scholars to be reminiscences or preserved traces of the Indo-Ary¬
ans and Iranians on their way to India and Iran from a western point of origin.

Misra’s contention is that, since none of these tribes brought Finno-Ugric Cauca¬
sian (or even Chinese or Korean) loanwords from these areas into the Avesta or Veda,
they could not have been coming from these areas to Iran or India. If, on the other
hand, the reverse were to be considered—that they were traveling from India and east¬
ern Iran/Afghanistan to these other areas—then the lacunae would be much better ac¬
counted for. In such a scenario, these other language families could have logically bor¬
rowed from these emigrating Indo-Aryans and Iranians before subsuming them. Other
Indian scholars have responded to such evidence in similar ways. Let us see if the Avesta
can discount these claims or throw any light on the Indo-lranian homeland.

The Avestan Evidence

The oldest parts of the Avesta, which is the body of texts preserving the ancient canon
of the Iranian Zarathustran tradition, is linguistically and culturally very close to the
material preserved in the Rgveda. Zarathustra preserved some of the cultural Indo-lra¬
nian features common to the Rgveda and developed, reformed, or rejected others

Linguistic Evidence from outside of India

131

(Humbach 1991, 2). 19 Operating within the same social and religious milieu that is
reflected in the Rgveda, Zarathustra’s teachings are to some extent defined vis-a-vis this
milieu. Both cultures, for example, place enormous economic importance on cattle.
However, unlike in the Vedic ethos, where cattle raiding is glorified as a heroic under¬
taking, Zarathustra condemns such exploits as wicked and selfish. Likewise, he opposes
the worship of devas (who are the chief benefactors in the later Vedic tradition), demon¬
izes them, and supplants them in terms of righteousness with the benevolent Ahuras
(Vedic asuras, who are in turn depicted as malevolent entities in most of the Vedic tradi¬
tion). 20 There seems to be economic and religious interaction and perhaps rivalry operat¬
ing here, which justifies scholars in placing the Vedic and Avestan worlds in close chro¬
nological, geographic, and cultural proximity to each other not far removed from a joint
Indo-Iranian period. If the evidence preserved in the Iranian tradition, the Avesta, con¬
tradicts the hypothesis of an Indo-Aryan linguistic community indigenous to India, then
one need waste no further time speculating on how the other members of the Indo-
European family might be accounted for in the Indigenous Aryan scenario.

Linguistically, the oldest sections of the Avesta are almost identical to the language
of the Rgveda. This oldest part is called the Yasna (Vedic yajna), which is dte principal
liturgical work of the sacred canon and accounts for about a third of its bulk. The nucleus
of the Yasna consists of the five Gathas (Vedic gatha ‘song’), a collection of seventeen
hymns (each Gatha containing from one to seven hymns), which are the only authentic
literary heritage left to posterity by Zarathustra himself (Humbach 1991, 3). Along with
a few other texts in the corpus, these Gathas contain the most archaic language of the
Avesta. The language displays the same richness as Vedic in its verbal inflection, con¬
tains almost identical morphology, exhibits the same processes of word formation, and
consists, for the most part, of words that have cognate forms in Vedic that are distin¬
guished by minor phonetic changes (Kanga and Sontakke 1962, xx). 21 The proximity of
this language to Vedic is so remarkable that whole Avestan passages can be transformed
into Vedic simply by making these minor phonetic correspondances (e.g., Sanskrit s =
Arestan ft). 22 With such close economic, religious, and linguistic overlap, the two texts
are considered to be very similar literary offshoots of a not-too-distant, proto-Indo-Ira-
nian period. This is significant because if Zarathustra’s date can be determined with
any degree of accuracy, it might serve as an anchor to secure the date of the compilation
of the Veda in close chronological proximity, since, as noted earlier, the oldest part of
the Avesta is attributed to Zarathustra.

Zarathustra’s date, unfortunately, is far from certain. Previously, a sixth century b.c.e.
date based on Greek sources was accepted by many scholars, but this has now been
completely discarded by present-day specialists in the field. Two dates for the prophet
were current in Greek sources: 5000 years before the Trojan War, that is, 6000 b.c.e.,
and 258 years before Alexander—the sixth century b.c.e. date. This more modest date
has been shown to be completely fictitious, but it initially gained wide acceptance be¬
cause it seems to have been adopted by the later Zarathustran scholastics themselves in
the Pahlavi books (Boyce 1992, 20). Since there is no specific historical information in
the tradition itself, once this chronological anchor had been unfastened, dating the Avesta
can only be based on the same conjectural suppositions that characterize attempts to
date the Veda.

The lower date for the Avesta, as with the Veda, is comparatively easy to establish.

132 The Quest for the Origin of Vedic Culture

Although there is much that is very arbitrary in the assignment of figures to genealo¬
gies, Boyce (1992, 29) claims a lower date of no later than 1100 b.c.e. for the Gathas
on the basis of the lineages, and so on, recorded in the texts. 23 As for the higher date,
Boyce has less to work with. Since, like the Rgveda, the Avesta is well acquainted with
chariots, and since the earliest remains of a spoked-wheel chariot dated to before 1600
B.C.E. have been found in the Sintashta Cemetery in the Inner Asian steppes (from
where Boyce is assuming the Iranians entered Iran), she tentatively proposes an upper
limit of 1500 b.c.e. for the Gathas (44). Gnoli (1980, 159), who is also supportive of
an earlier date (end of the second, beginning of the first millennium b.c.e.) along many
of the same lines as Boyce, points out that the oldest Avestan texts reflect a milieu that
is certainly much different than that of the Medes and the Achaemenians. Significantly,
the texts make no mention of urban centers, or even of geographic regions in the west
of Iran. Such features all indicate a decidedly prehistoric period. When we consider
that these tentative dates are bolstered by coordination with the generally accepted date
for the composition of the Vedic hymns (Boyce 1992, 29) and the movements of the
Indo-Aryans (Boyce 1992, 41)—both of which, from the perspective of this study, are
under reconsideration—it becomes clear that the Avestan evidence, rather than helping
to secure the Vedas temporally, simply becomes an extension of the same chronological
problem.

Geographically, the Avesta has little to offer the quest for the homeland of the Indo-
Aryans speakers—with one very important exception. In sharp contradistinction to the
lack of any clear reference in the Vedic tradition to an outside origin, the Avesta does
preserve explicit mention of an airiiansm vaejo, the legendary homeland of the Aryans
and of Zarathustra himself. 24 The descriptions of this place, despite the fact that “it is
revealed that Ohrmazd made [it] to be better than the other places and regions,” speak
of severe climatic conditions (Humbach 1991, 35). 25 Gnoli (1980, 130) situates the
airiiansm vaejo in the Hindu Kush because all the identifiable geographic references in
the Avesta are of eastern Iran, south central Asia and, Afghanistan, with an eastern
boundary formed by the Indus. There is no mention of any place north of the Sir Darya
(the ancient Jaxartes), 26 nor of any western Iranian place (Boyce 1992, 3).

Skjaervo (1995) finds the identity of the airiianom vaejo to be insoluble, but finds it
possible that it might have changed as the tribes moved around. He reiterates the noto¬
rious difficulties involved in using the Avesta as a source for the early history of the
Iranians and cautions against circularity, since, if “we use archaeology and history to
date the Avesta, we cannot turn around and use the Avesta to date the same archaeo¬
logical and historical events, and vice versa” (158). While skeptical of attempts to pin¬
point the exact locus of their composition, he concurs that the internal evidence of the
Avestan texts makes it impossible to avoid the conclusion that they were composed in
northeastern Iran and traveled from there to the south and southwest (165-166). Witzel
(forthcoming, b) undertakes a thorough analysis of all the philological, linguistic, envi¬
ronmental, and climactic pointers in the texts and feels the Avestan homeland points to
the central Afghan highlands including, perhaps, areas north and south of the Hindukush.

So the geographical boundaries of the Avesta are approximately coterminus with the
western part of the area under consideration by Indigenous Aryanists (and the favored
homeland of nineteenth-century scholars). Of course, all this could be a reference to a
secondary Indo-Iranian homeland, and not the primary Indo-European one, but, like

Linguistic Evidence from outside of India 133

the Veda, the Avesta preserves no clear memory of an overland trek from the Caspian
Sea. However, the Hapta Hendu (Vedic Sapta Sindhu or Panjab) is known, as is the
Harahvaiti (Vedic Sarasvatl), 27 which denotes an Iranian river. Boyce, in harmony with
most Indo-Europeanists, believes that the oblique geographic indicators in the Avesta
refer to the Inner Asian steppes, but, regrettably, there is little that is of real help in
locating the homeland of the joint Indo-lranian speakers unless one is prepared to ap¬
ply the data to a homeland already prefigured on other grounds. 28 Nor is there any
reason for Indigenous Aryanists to disagree that the Avestan airiiansm vaejo could have
been Afghanistan and the Hindu Kush, as these scholars suggest.

There are also identical names of rivers common to both Iran and India, such as the
Iranian Harahvaiti and Haroyu, which correspond to the Indian Sarasvatl and Sarayu
(Sanskrit 5 = Iranian h). In and of themselves, all that can be said of this data is that
these names could have been either transferred by incoming Indo-Aryan tribes from
Iranian rivers to Indian ones, as is generally assumed, or by outgoing Indo-lranian tribes
from Indian rivers (although any transfer from Iran to India must have occurred before
Iranian developed the h phoneme, since s can become h but never vice versa). Scholars
have conventionally interpreted these transferals as evidence of the movement of the
Indo-Aryans toward India from the Caspian Sea area via Iran. Burrow, (1973a) for ex¬
ample, considers that the Vedic name Sarasvatl belonged originally to the river in Iran
that it refers to in the Avesta and was later imported into India by the incoming Indo-
Aryans. Another set of Indo-lranian river terms river—Sanskrit ‘rasa’, Avestan ‘Ragha —
has been identified with cognates in Russia, where ‘Rosa’ is a frequent river name, which
to some is an indication of the steppe origins of the Indo-Aryans. A number of Baltic
river names have the form ‘Indus’, ‘Indura’, ‘Indra’, and so on, which are explainable
by comparison with Sanskrit indu ‘drop’ (Mallory 1975, 169). These hydromic etymolo¬
gies have been accepted as signs of Indo-Aryans (or Indo-Iranians or Iranians) moving
across Asia toward their historic seats in India and/or Iran.

Indigenous Aryanists maintain that they could all just as easily be signs of tribes
emigrating from the Indian subcontinent and its environs toward Iran and the North¬
west. Even Max Muller (1875) considered that the “Zoroastrians were a colony from
Northern India. ... A schism took place and [they] migrated westward to Arachosia
and Persia. . . . They gave to the new cities and to the rivers along which they settled,
the names of cities and rivers familiar to them, and reminding them of the localities
which they had left” (248). More recently, Erdosy (1989) has also noted that “it would
be just as plausible to assume that Saraswati was a Sanskrit term indigenous to India
and was later imported by the speakers of Avestan into Iran. The fact that the Zend
Avesta is aware of areas outside the Iranian plateau while the Rigveda is ignorant of
anything west of the Indus basin would certainly support such an assertion” (42). Cer¬
tainly, the Indus Valley, which Indigenous Aryans, as we shall see in the next chapter,
consider to have been Indo-Aryan, formed a trading and cultural zone with central Asia,
Afghanistan, and Baluchistan in the third millennium b.c.e. The Avesta, then, simply
deepens the mystery of Indo-Aryan origins.

Lubotsky (forthcoming) has identified a number of words shared between the Indo-
Aryan and Iranian branches that are not evidenced in any of the other Indo-European
languages. Some of these are inherited from Indo-European, but others exhibit phono¬
logical or morphological features that make them conspicuous as loans. Lubotsky finds

134 The Quest for the Origin of Vedic Culture

that the structure of these loans in Indo-Iranian is similar to the structure of the loans
in the Rgveda found by Kuiper, suggesting that they were borrowed from the same lan¬
guage (or, at any rate, two dialects of the same language). In Lubotsky’s view, “in order
to account for this fact, we are bound to assume that the language of the original popu¬
lation of the towns of Central Asia, where Indo-Iranians must have arrived in the sec¬
ond millennium bce, on the one hand, and the language spoken in the Punjab, the
homeland of the Indo-Aryans, on the other, were intimately related” (4). Since there are
some irregularities in the Iranian correspondances, Lubotsky further infers that the Indo-
Aryans formed the vanguard of the Indo-Iranian collective heading southeast and were
thus the first to come into contact with the tribes who spoke foreign languages. After
adopting the loan words from these tribes, they passed them on to the Iranians, who
adjusted the phonemes somewhat, thus producing these irregularities.

As a side note, if the subtratum language of Central Asia and drat of the Punjab
were the same, then the question arises as to why Iranian was not affected by at least
some of the same non-Indo-European phonological and morphological features unique
to Vedic and Sanskrit that have generally been assigned to the Punjab substratum, as
was discussed in chapter 5. Also of interest is the fact that Lubotsky is surprised to find
that many of the loans into Finno-Ugric are only attested in Sanskrit, and not Iranian—
an observation that would doubtlessly be of use to Misra’s thesis as outlined here.
Lubotsky holds that this could be because the Indo-Aryans were the vanguard of the
Indo-Iranians, but one might wonder why die Iranians in the rearguard did not like¬
wise impart loans.

Lubotsky (forthcoming) notes that the landscape of Indo-Iranians must have been
quite similar to their original homeland since there are no new terms for plants or other
items of the environment. There are loans for new animals such as the camel, as well
as for irrigation, elaborate architecture (including permanent houses), clothing, and hair
styles. There are few terms for agriculture, reinforcing the idea of their nomadic lifestyle,
and there is a paucity of terms for military technology, which underscores Aryan mili¬
tary supremecy. All of this “is a strong confirmation of the traditional theory that the
Indo-Iranians came from the North. . . . The Indo-Aryans formed the vanguard of the
Indo-Iranian movement and first came in contact widr the original inhabitants of the Central
Asian towns. Then, presumably under pressure of the Iranians, who were pushing from
behind, the Indo-Aryans moved further south-east and south-west, whereas the Iranians
remained in Central Asia and later spread to the Iranian plateau” (5).

While this is a perfectly satisfactory interpretation of the data, these foreign terms in
Indo-Iranian are subject to the same possibilities as the substratum elements discussed
in chapter 5: How can one dismiss the possibility that it was the foreign terms that were
entering into an Indo-Iranian speaking language group already situated in East Iran,
Afghanistan, and the Punjab, rather than the opposite—the Indo-Iranian speakers im¬
migrating into the territory of a preexisting indigenous language group situated in the
same area? A good number of the terms—‘bad,’ ‘smell,’ ‘head hair,’ ‘milk,’ ‘belly,’ ‘spit,’
‘tail,’ ‘heap,’ ‘penis,’ etc.—are surely items that exist in any language. If the Indo-Irani¬
ans borrowed such terms, it is not because they did not already possess equivalents in
their own dialects. Terms often get replaced and new synonyms are added to a vocabu¬
lary as tribes interact with each other. As for the terms denoting new or unfamiliar items
such as garments, the camel, or bricks, how can we eliminate the possibility that these

Linguistic Evidence from outside of India 135

were not simply exotic, desireable, or useful foreign items that entered a language area
through trade or other means and that retained their foreign names?

Moreover, as has already been discussed, what mechanisms can linguistics provide
to determine whether such loans are due to substratum, superstratum, or adstratum?
Lubotsky (forthcoming) himself notes that “I use the term substratum to refer to any
donor language, without implying sociological differences in its status, so that ‘substra¬
tum’ may refer to an adstratum or even superstratum” (1). These are all issues that need
to be resolved before determing that such terms reflect preexisting substratum languages
that can be used to chart the movements of the Indo-Aryans.

The Mitanni Treaties

Another set of data crucial to the Indo-Aryan saga, and one of the few happy occasions
when archaeology and linguistics coincide unambiguously, is the inscriptional evidence
demonstrating die presence of Indo-Aryans throughout the Near East in the fourteenth
century b.c.e. By the turn of the twentieth century, Indologists in the West had long ceased
to contest the picture of an overland trek of the Indo-Aryans from a European or Caspian
Sea homeland across the Asian steppes into India. It came as a great surprise to learn
about these Indo-Aryans that had suddenly come to light as a result of Hugo Winckler’s
excavations in Bogazkoy, Asia Minor, during die summer of 1907. These Indo-Aryans
were not just misplaced wondering nomads who had happened to take a premature turn
at the wrong steppe, but rulers of the Mitanni kingdom in north Mesopotamia and rulers
of neighboring principalities as far as Syria and Palestine, How did they get to be there?

The principal texts that have generated the most interest among scholars interested
in this topic are the treaty between a Hittite king and a Mitanni king wherein the Vedic
gods Indara (Skt. Indra), Mitras(il) (Skt. Mitra), Nasatia(nna) (Skt. Nasatya), and
Hrumnass(il), (Skt. Varuna) are mentioned, and a treatise on horse training and upkeep

The Mitanni kingdom.

136 The Quest for the Origin of Vedic Culture

by a Mitannian called Kikkuli wherein technical terms related to horsemanship are in
Indo-Aryan. (For a comprehensive discussion of the Bogazkoy texts, see Starke 1995.)
The text suggests Indo-Aryan prominence in this field. Aryan traces also appear in Kassite
documents from the Babylon dynasty, such as Surias (Skt. Surya) and the war god Mariettas
(Skt. Marut, although this latter always occurs in the plural). In addition, over a hun¬
dred such names of local personalities have been produced by scholars. (See, for ex¬
ample, Dumont 1948, 56-63 and appendix.) The Indo-Aryan character of these is ac¬
cepted by most scholars. Moreover, they referred to kings and ruling elite spread out
over an area that included Syria and Palestine. 29 Although these rulers were Indo-Aryan,
the population was not, being mostly Hurrian. This suggests elite dominance of the
Mitanni state and a large area of neighboring territories by an intruding, martial, Indo-
Aryan class.

Jacobi, more concerned with the temporal as opposed to the geographic significance
of the findings, was delighted at the news. He and Tilak, as will be discussed in chap¬
ter 12, had utilized Vedic astronomical references to argue for a much older date for the
Rgveda than Western scholars had ever considered before, but both attempts had been
criticized by some of their peers. Urging restraint on the belligerent Tilak (who was
unbowed by the negative responses to his astronomical theories, and “who wished to
enter upon a campaign against all opponents”), Jacobi told him that “the discussion
would have no definite result unless excavations in ancient sites in India should bring
forth unmistakable evidence of the enormous antiquity of Indian civilization” (Jacobi
1909, 722). Learning of the findings at Bogazkoy, Jacobi felt somewhat vindicated that
at least the late date his colleagues were proposing for the Rgveda was now discredited,
since the Mitannians, whom Jacobi assumed had come from India, were already wor¬
shiping the Vedic gods in “full perfection” in the fourteenth century b.c.e. Obviously,
their worship must have been even earlier in India itself, so that he might consequently
“perhaps think that [his] chronological argument will yet survive” (726). It has been
suggested that some of the names go back to the seventeenth century b.c.e. (Dumont
1948, 63) or earlier still (Akhtar 1978-79, 66).

As might be expected, the treaties provoked considerable discussion for some time
over whether the names and terms recorded in them were Indo-Iranian, Iranian, or
Indo-Aryan. Paul Thieme’s study of 1960 concluded the matter to the satisfaction of
most scholars by demonstrating that the names were specifically Indo-Aryan. 30 Four
possibilities have been considered by scholars regarding how these Indo-Aryans ended
up in the Near East. First, they might have been the complete Indo-Aryan group that,
after its split with the Iranian members of the original Indo-Iranian collective, initially
took over the Mitanni kingdom and then proceeded eastward to colonize the North¬
west of the subcontinent. This possibility is not presently considered by most scholars.
Another alternative, to which most Western historians subscribe, is that these Aryans
were a segment of the Indo-Aryans (after the split with the Iranians) somewhere in north
Iran or central Asia who peeled off from the main group of Indo-Aryans who were
migrating east toward India. Leaving the larger body, they sought their fortunes in the
Near East, where, although successful, they eventually became subsumed by the local
population. A third possibility is that they were a part of the unseparated body of Indo-
Aryans who initially reached India but, sometime after arriving, bade farewell to their
kinsmen, retraced their steps, and headed back east. 31 Sten Konow (1921, 37), who

Linguistic Evidence from outside of India 137

agreed with Jacobi’s proposal that the Vedic culture in India must have had a much
higher antiquity than most scholars were willing to allow, argued for a version of this
third possibility. He proposed that although the Indo-Aryans were immigrants to India,
once they had established the Vedic culture on the subcontinent, some of them traveled
back out of India and into Mesopotamia. This chain of events would require assigning
the Rgveda a date considerably older than that of the Mitanni treaties.

The fourth possibility, supported by Jacobi and Pargiter in the West, and often viewed
as the least complicated one by Indigenous Aryanists, is that these Indo-Aryans could
have been Vedic-speaking tribes from the Indian subcontinent who left their homeland
in the Punjab for the Near East, bringing their favorite gods with them: “[The Mitanni
evidence] can either mean that the Aryans were on their trek to India from some up¬
land in the north or the Indo-Aryan culture had already expanded from India as far as
Asia Minor” (Vidyarthi 1970, 33); “Did the worshippers of Indra go from an earlier
home in the Indus valley to Asia Minor, or was the process just the reverse of this?”
(Majumdar 1951, 25). There is nothing in the Near Eastern documents themselves that
militates against either the third or the fourth possibility. On the contrary, as will be
discussed more thoroughly in chapter 10, archaeologists point out that there is not a
single cultural element of central Asian, eastern European, or Caucasian origin in the
archaeological culture of the Mitannian area (Brentjes 1981, 146). This is a significant
obstacle to those proposing versions of the first or second possibilities.

In contrast to this lacuna, Brentjes draws attention to the peacock element that recurs
in Mitannian culture and art in various forms (to be eventually inherited by the Irani¬
ans), a motif that could well have come from India, the habitat of the peacock. Since
this motif is definitely evidenced in the Near East from before 1600 b.c.e., and quite
likely from before 2100 b.c.e., Brentjes (1981) argues that the Indo-Aryans must have
been settled in the Near East and in contact with India from well before 1600 b.c.e . 32
The corollary of this is that the Indo-Aryans “could not be part of the Andronovo cul¬
ture [a culture dated around 1650-1600 b.c.e. with which they are usually associated],
but should have come to Iran centuries before, at the time when the Hittites came to
Anatolia” (147).

Satya Swarup Misra (1992) again presents some linguistic comments in support of
this greater chronology. 33 He argues that many of the linguistic features in the Anatolian
documents are much later than Vedic but identical to the forms found in Middle
Indo-Aryan. 34 These were also noticed by Kenneth Norman (1985, 280). 35 Hodge
(1981) also draws attention to satta ‘seven’, which is the Prakrit form of Sanskrit sata,
and remarks that the inscriptions show a Prakritic form of Sanskrit a thousand years
before such forms are known in India itself on inscriptions. These observations fit
comfortably with the proposal that the Near Eastern kings could have left the Indian
subcontinent after the early Vedic period, bringing post-Vedic, Indo-Aryan linguistic
forms with them. The most drastic corollary of such a claim, as Jacobi noted, would
be a major reevaluation of the dating of the Rgveda, which must have considerably
predated the appearance of the Near Eastern texts in 1600 b.c.e. if these do, indeed,
represent a diachronically later, as opposed to a synchronically contemporaneous, or
dialectal, form of Indo-Aryan.

There has been some controversy over the numeral aika ‘one’, which some linguists
hold could be from Indo-Iranian which has the form aika (e.g. D’iakonov 1985, 158),

138 The Quest for the Origin of Vedic Culture

rather than from Indo-Aryan. However, (the Iranian term for ‘one’ is aim, but it is quite
possible that Indo-Iranian could have had two terms for ‘one’ since eva survived in
Sanskrit as ‘alone’) suggesting that this term could indeed be Indo-Aryan. 36 Misra (1992),
and others we can recall, hold that the loanwords in Finno-Ugric were Indo-Aryan, which
if correct would indicate that the Indo-Iranians had split into Iranians and Indo-Aryans
long before the Mitanni treatise. However, Norman (1985, 280), in contrast to Misra,
accommodates the possible Middle Indo-Aryan linguistic features in the treatise with
the commonly held opinion regarding Indo-Aryan origins by proposing that Middle
Indo-Aryan dialectal variations were already in existence within the Indo-Aryan com¬
munity as they were en route to India. Of relevance here is that scholars (e.g. Elizarenkova
1989) have noted certain Middle Indo-Aryan forms present in Vedic but absent in San¬
skrit, which lends support to this possibility of Middle Indo-Aryan forms being extant
as optional or dialectal forms during the Vedic period itself. Accordingly, since Misra’s
observations can be interpreted as indicating that these MIA forms were already extant
in the Indo-Aryan language on its way to India, they do not prove a greater antiquity of
the Veda. But Misra’s claim remains a possibility nonetheless. In fact, all of the various
arguments presented by Indigenous Aryanists so far reject the 1200 b.c.e. date com¬
monly assigned to the Rgveda. This is a most crucial issue, and one to which I will turn
in chapter 12.

Conclusion

In conclusion, then, the recurrent theme of this work has been to show how different
perspectives and assumptions result in different interpretations of the evidence and
produce different conclusions. From post-Proto-Indo-European to the historical period
(as known from Greek sources), we find traces of Indo-Iranian, Indo-Aryan, and/or
Iranian speakers spread out over a vast area stretching from North India across the
Middle East, central Asia, the Caucasus, and up to the Caspian Sea. Indigenous Aryanists
believe nothing in any of this evidence can deny the possibility that the Indo-Aryan and
Iranian tribes might have been emigrating from a northwest South Asian/Afghanistan
homeland toward the Northwest as opposed to immigrating from the Northwest to the
Southeast.

As a side note, one reference in particular is repeatedly produced from the Puranas
as evidence of a large emigration from Gandhara, Afghanistan, to the northern regions.
The narrative is situated in the time of Mandhatr, who drove the Druhyu king Angara
out of the Punjab. Pargiter ([1922] 1979) notes that the next Druhya king, Gandhara,
retired to the Northwest and gave his name to the Gandhara country (which survives to
the present day in the name Kandahar in Afghanistan). The last king in the Druhyu
lineage is Pracetas, whose hundred sons take shelter in the regions north of Afghani¬
stan ‘udicim dis'am asritah’and become mlecchas. 77 The Puranas make no further refer¬
ence to the Druhyu dynasty after this. 38 The more enthusiastic see this as “evidence of
the migration of Indo-Europeans from India to Europe via Central Asia” (Talageri 1993,

367).

Most scholars, of course, will respond to such a claim by pointing out that the Puranas
are generally considered to have been written much later than the terminus post quern

Linguistic Evidence from outside of India 139

date for the Indo-European dispersal. This dispersal must have occurred before the earliest
recordable evidence for distinct, already separated IE languages, such as the Hittite and
Mitanni documents, which are datable to the first half of the second millennium b.c.e.
Scholars do acknowledge that much material in the Puranas does go back to the Vedic
age—indeed, Rocher, who posits an ur-Purana, declines to even attempt to date them—
and although the Druhyus are certainly mentioned in the Rgveda as one of the tribes
fighting in the dasarajna war, scholars are hardly likely to assign an earlier date to some
of the contents of the Puranas than the date they have assigned to the Rgveda of 1500—
1200 b.c.e. Not surprisingly, the Indigenous Aryan school disagrees completely. For
them, the assignment of such dates is rather arbitrary, and most are open to the possi¬
bility of a Veda as old as the third, or even fourth, millennium b.c.e. coexisting with
strands from the contents of the material recorded in the Puranas.

Be all that as it may, it certainly cannot be stated that all historical movements con¬
nected with the Northwest of the subcontinent have been from west to east, as some
scholars imply when comparing the Indo-Aryan migrations with those known to his¬
tory such as the Greeks, Kusanas, Sakas, Moghuls, and others. Scholars of the nine¬
teenth century protested against the opposite tendency of their time, which was consid¬
ering people’s movements to have always been from east to west—Michelet’s ex Oriente
lux. In addition to the earlier reference from the Puranas, Indian scholars have often
pointed out that, in later times, the Gypsies emigrated from India to the West, as did
Indian Buddhists to the Northwest, influencing significant areas of central Asia (and,
of course, China and the East). Hock (1993, 82) draws attention to other Indo-Aryan
groups from the subcontinent who have followed a northern trajectory: Gandhari or
Niya Prakrit in early medieval Khotan and farther east; modern Dumaki in northwest¬
ern South Asia; and the Parya who came to modern Uzbekistan via Afghanistan. Ear¬
lier still, Indus seals found their way to Bahrain, Mesopotamia, and central Asia (al¬
though this is likely to have been through trade rather than emigration). Of course, the
historic situation need have nothing to do with the protohistoric one, but the point
sometimes made by Indigenous Aryanists, of course, is that South Asians have not been
reluctant to emigrate to the Northwest in the modern, historic, and protohistoric peri¬
ods (to which even the modern-day South Asian diaspora attests), so, by comparison,
there is no a priori reason to suppose that this could not have happened in prehistory.

But such suppositions have little to do with serious arguments. The ultimate point,
from the perspective of the present discussion, is not that the evidence in this chapter
in any way suggests a South Asian homeland. The point is that it merely cannot ex¬
clude it as one possibility among a number of others: there is little in the matter of
loanwords outside the Indian subcontinent that might convincingly persuade Indigenous
Aryanists to change their view. Much of the evidence is malleable and has been recon¬
figured by scholars with different perspectives and presuppositions such as Misra. Here,
too, compelling proof of Indo-Aryan origins eludes us.

The Viability of a South Asian Homeland

Since the principle that cognate languages stem from some kind of a protoform has yet
to be refuted, as has the postulate that protolanguages must have been spoken in some
kind of a reasonably delimited geographic area, there seem to be only three (or four)
options that could account for the connection of the Indo-European languages as a fam¬
ily. Either the Indo-Aryan languages came into India from outside or, if it is to be claimed
that the Indo-Aryan languages are indigenous to India, the corollary must be that the
other Indo-European languages left from India to their historically known destinations.
The third alternative is that there was a very large surface area stretching from the
Northwest of the subcontinent to the Caspian Sea wherein related, but not necessarily
homogeneous, Indo-European languages were spoken. Trubetzkoy (1939) offered a fourth
proposal, that Indo-European was a language created by the creolization of several dif¬
ferent languages in contact. 1

The third proposal is attractive in a “politically correct” sort of way in that it by¬
passes the need for subscribing to either an immigration into or an emigration out from
India; indeed, K. D. Sethna (1992, 75) has settled on this very solution, as has Kenoyer
(1997). However, linguists are likely to object that such an area is too vast to account for
the morphological, lexical, and other shared features of the Indo-European languages,
which would require a much more compact and limited geographic origin. Such a vast
geographic area of origin for a language family has never been attested in linguistics—
the Romance languages, as a point of comparison, originated in a limited area around
Rome, although they spread all over southern Europe. Unfortunately, the minute one
tries to further narrow this vast Indo-European-speaking area, one enters the quagmire
of speculation and disagreement that has been characteristic of the Indo-European home¬
land quest since its inception. What Western scholars have not been aware of until very

140

The Viability of a South Asian Homeland 141

recently is that some Indian scholars have utilized the same linguistic evidence used in
debates in Western academic circles to argue that even India cannot be excluded a priori
from being a possible homeland candidate.

This does not mean all Indigenous Aryanists believe that India was factually the
homeland of all the Indo-Europeans. Of course there are certainly those who, perceiv¬
ing the fallacies in many of the theories being promoted by their Western colleagues,
nonetheless attempt to utilize similar methods and logic to promote India as a home¬
land. Perhaps this is understandable after being subjected to two centuries of unbridled
European intellectual hegemony on the Indo-European homeland problem. But clearly
an Indian homeland theory is as open to the same type of criticism that Indigenous
Aryans have vented on other homeland theories. Most scholars simply reject the whole
endeavor as irremediably inconclusive, at best, and “a farrago of linguistic speculations,”
at worst. The more careful members of the Indigenous Aryan school, at least, simply
recognize that all that can be factually determined with the evidence available at present
is that “the Indo-Europeans were located in the lndus-Sarasvati valleys, Northern Iran,
and Southern Russia” (Kak 1994, 192). 2 From this perspective, if the shared morpho¬
logical and other similarities mandate that the Indo-Europeans had to come from a more
compact area, that is, from one side of this large Indo-European-speaking expanse, most
Indigenous Aryanists see no reason that it has to be the western side: “We can as well
carry on with the findings of linguistics on the basis that India was the original home”
(Pusalker 1950, 115). In other words, by arguing that India could be the Indo-European
homeland, the more cautious scholars among the Indigenous Aryanists are demonstrat¬
ing the inadequacy of the linguistic method in pinpointing any homeland at all, rather
than seriously promoting India as such.

Factually, however, the fact that so few scholars have really taken it upon themselves
to engage the linguistic issues is the Achilles’ heel of the Indigenous Aryan school. The
Indo-Aryan invasion theory was originally constructed on linguistic grounds, and it can
be effectively and convincingly dismantled only by confronting its linguistic infrastruc¬
ture. Even Indian archaeologists who have been the most stalwart supporters of Aryan
invasions or migrations defer to the linguistic evidence:

It has been repeatedly denied that there was any Aryan movement into India. . . . [A]r-
chaeology has not “proved” the intrusion of the Indo-Aryan languages into South Asia
. . . [but] equally . . . absence of such “proof’ is not tantamount to refutation of the theory
of linguistics. . . . [I]t appears as if archaeological data and methods are not appropriate
tools with which we may study the problem. From historical linguistics we learn that speakers
of the earliest/ancestral Indo-European languages originally lived in a common home¬
land in Eurasia and subsequently . . . dispersed . . . eastward into Iran and India. (Ratnagar
1996a, 1; my italics)

As we shall see in the following chapters, Soudr Asian archaeologists have indeed re¬
peatedly insisted that the archaeological record shows no sign of Indo-Aryan incursions
and that, therefore, the Aryan invasion theory is at best a linguistic issue; and, at worst,
it is simply a figment of nineteenth-century European imagination. As a point of com¬
parison, similar arguments were made in 1992 in a series of essays by a Western ar¬
chaeologist, Alexander Hausler (Lehmann 1993). Hausler found that the archaeological
evidence in central Europe showed continuous linear development, with no marked
external influences. This caused him to reject Gimbutas’s three-stage invasion hypoth-

142 The Quest for the Origin of Vedic Culture

esis from the Russian steppes, as well as the whole notion of an Ursprache “proto¬
language.” However, as Lehmann curtly points out, “He has no proposal for the source
of the Indo-European language or languages. . . .For Hausler. . .the notion of an
Ursprache is obsolete. Yet on the basis of linguistic study we must assume that soci¬
eties maintained common languages that, like languages today, were open to change.
. . .Accordingly we posit proto-languages” (285-286). Archaeologists, whether West¬
ern or Indian, cannot simply dismiss, or ignore, linguistic reality.

Center of Origin Method

As noted previously, the first serious attempt I have found that challenges the prevail¬
ing ideas regarding the Aryan invasions by utilizing linguistic, as opposed to philologi¬
cal, data was by Lachhmi Dhar in 1930. The then commonly accepted status quo, as
mentioned in chapter 1, had been established by Robert Latham, an ethnologist, who
was the first to challenge the idea of an Asian homeland. Latham had proposed that
languages are analogous to biological species: the geographic center of origin of a spe¬
cies exhibits the greatest variety of features. The homeland of the Indo-Europeans there¬
fore must be found wherever the greater variety of language forms were evidenced, that
is, in or near Europe. The Indo-Iranian languages, in contrast, being more homoge¬
neous, were more peripheral to the area of greatest variety and therefore must have been
peripheral to the homeland. More recently, Dyen (1965) has articulated this principle
on similar grounds:

The strongest hypothesis regarding the homeland of peoples speaking languages belong¬
ing to the same language is one that assigns this homeland to the area in which the ge¬
netically most diverse members of the family are to be found. It is reasonable that the
whole of a large number of groups of people is not likely to migrate as a collection of
distinct groups. If then we find the most diverse collection of languages belonging to the
same family in a particular area, there is a prima fade argument that the languages grew
different in that area rather then elsewhere. (15)

In other words, it is more probable that one or two groups moved out from a geographic
maitrix that had become linguistically heterogeneous, than that many linguistically dis¬
tinct groups moved out from a linguistically homogeneous area.

Although this line of reasoning is certainly more sophisticated than the proposals to
simply situate the homeland in the exact geographic center of the whole Indo-European-
language-speaking zone, various objections were raised, most notably that the existence
of the Tocharian language in the eastern group disrupted its homogeneity somewhat.
At the time of writing, there is still no consensus on proto-Bangani, which Zoller has
claimed to be an Indo-European language in India itself that contains very archaic Proto-
Indo-European features that would significantly demarcate it from Inclo-Aryan. In ad¬
dition, the center of gravity argument, which is no longer accepted by most linguists,
ignores possibly more important factors that will be discussed later. 3

In 1930, in response to Latham’s hypothesis, Lachhmi Dhar provided a different
explanation for this greater linguistic diversification in the western Indo-European lan¬
guages of Europe. Dhar argued that the European side of the family, upon leaving its

The Viability of a South Asian Homeland 143

original homeland, which he situates in the Himalayas, had to impose itself on the
indigenous, pre-Indo-European languages known to have existed in Europe (such as
Basque, Etruscan, Iberian, and Pictish) and thereby absorbed more foreign linguistic
elements. 4 This resulted in the greater linguistic variety perceivable in Europe. The Indo-
Iranian languages, in contrast, developing organically from Proto-Indo-European, which
was native to India, absorbed no such alien influences and thereby remained relatively
homogeneous and conservative. Dhar’s position invokes a linguistic principle based on
the conservation principle. This holds that the area of least linguistic change is indica¬
tive of a language’s point of origin, since that area has been the least affected by sub¬
strate interference. The very characteristics used by Western scholars to postulate a
European homeland were used, by Dhar, as evidence that this homeland could certainly
not have been in Europe, and the disqualification assigned to Indo-Iranian, he reconstrued
as its qualification. Actually, Dhar’s line of argument has a history in Western debates
in the Indo-European homeland, (e.g., Feist, 1932; Pissani, 1974).

Dhar (1930) went on to argue that “ancient Sanskrit possesses the greatest num¬
ber of roots and words and the greatest variety of grammatical forms, belonging to
the Aryan mother tongue, when compared with all the other Aryan languages in the
world” (59). Here, Dhar is referring to the long-established position of his time: “The
Sanscrit is the language which has retained the most primeval form and has adhered
the most tenaciously to that parent ground. . . . [It] has preserved a great number of
roots which have been lost in the other languages (Weber 1857, 6). Most particu¬
larly, Dhar singled out the accent as the specific feature of the Vedic language that is
homogeneous with Proto-Indo-European, since “every student of comparative philol¬
ogy knows that the Vedic Sanskrit preserves most faithfully the accent of the Aryan
mother-tongue, although, quite naturally, traces of the original Aryan accent are also
noticeable in some other Aryan languages such as the Greek and Pre-Germanic lan¬
guages etc.” (45). Dhar wonders:

How is it that as the speakers of the language traveled all their way from Europe to Asia and
then finally settled in India, they were able to retain in India alone of all countries—their
final destination which they must have reached after a course of several centuries—almost
exactly the same accent on words which their European forefathers used to possess centu¬
ries before in their forest-home in Europe or their Asiatic fathers on the table-land of Asia
away from India, but which their bretheren in different countries . . . could not preserve. . . .
Nor can the Aryans be supposed to have traveled through an ethnic vacuum as they started
their journey from Europe or outer-Asia and traveled across thousands of miles of land
before they could reach India, and escaped the influence of alien speech habits on their
language. . . . [on the contrary, it was] the Aryas in their journeyings in Europe and Asia
outside India [who] could not avoid . . . [being] overwhelmingly swamped by foreign people.

. . . thus the ethnic disturbances have disturbed the original Aryan [Indo-European] accent
in Outer-Asia or Europe. . . . the continuity of the original Aryan accent in ancient India
implies the unbroken geographical and ethnic continuity of the Aryan race from the most
primitive times in India. . . . Thus the home of the primitive Aryan language can be located
round the home of the Vedic speakers who possessed almost exactly the same word or
sentence-accent as their Aryan [Indo-European] fathers did. (47—51 ) 5

It is true that the Vedic accent “reflects most faithfully the original Indo-European ac¬
centual system” (Halle 1977, 210; see also Kiparsky, 1977), but, of course, Dhar’s rea-

144 The Quest for the Origin of Vedic Culture

soning can be countered by arguing that Vedic retained the Indo-European accent be¬
cause it was a sacerdotal language, which artificially preserved forms that would other¬
wise have evolved in a normal spoken language; there may be other factors involved in
the conservation of archaic features apart from absence of substratum. Nonetheless, Dhar
made a serious effort to address the full range of linguistic data available to him, and
some of his arguments were as well-reasoned as anything that was on offer by some of
his Western colleagues of the day.

Other Indian scholars have repeatedly raised the same objections with regard to other
archaic features in Indo-Aryan:

The Vedic Sanskrit has the largest number of vocables found in the Aryan languages. . . .
if the pre-Vedic Aryan language was spoken in different parts of Europe and Asia where
the Aryans had settled, . . . how is it that only a few vocables are left in the . . . speech of
those parts, while the largest number of them is found in the distant places of ultimate
settlement and racial admixture in India? On the contrary, this disparity can easily be
explained if the pre-Vedic was the language of the homeland of the Aryans and the other
Aryan languages came into existence as a result of the contact between migrating Aryans
and non-Aryan elements outside India and Persia. (Majumdar 1977, 216)

The same arguments recur almost verbatim in numerous publications (Luniya 1978,
71; Pillai 1988, 78) and must have been widely circulating in India—perhaps because
some nineteenth-century Western scholars who were partial to a Bactrian homeland
had made the same case: “The nearer a language is to its primary centre, the less alter¬
ation we are likely to find in it. Now of all the Aryan dialects, Sanskrit and Zend may,
on the whole, be considered to have changed the least” (Sayce 1880, 122).

Although questionable, such arguments are interesting if only because the same logic
has been used repeatedly regarding Lithuanian as well as, more recently, Anatolian.
Both of these languages have also preserved very archaic Indo-European features, caus¬
ing them to be promoted as being situated in, or adjacent to, a postulated homeland.
The archaisms in Lithuanian were first accounted for by Latham as being due to this
language’s being spoken in the vicinity of the original homeland north of die Black Sea,
where it was not subject to alien substratum influence. 6 More recendy, Kortlandt (1990),
while promoting the eastern Ukraine as the best candidate for the original Indo-Euro¬
pean homeland, holds that the language of the Indo-Europeans who remained in this
area after the migrations eventually evolved into Balto-Slavic. Accordingly, “the decep¬
tively archaic character of the Lithuanian language may be compared to the calm eye of
a cyclone” (136-137). Along the same lines, as additional support for their Anatolian
homeland, Gamkrelidze and Ivanov state that “proto-Anatolian moved a relatively small
distance from the proto-Indo-European homeland. This explains the extreme archaism
of the Anatolian languages” (1995, 790). Dhar, then, adopted a line of reasoning that
is well used in such debates.

It seems clear that linguists have presented many factors as being involved in the
mechanics of language change, which vary according to time, place, and individual cases.
Few, nowadays, will accept overly simplistic explanations such as that either variegation
and innovation or homogeneity and archaicness automatically indicate autochthony.
Conservatism has indeed been demonstrated as corresponding with indigenousness in
a number of case studies (e.g., Ross 1991), and foreign substrata influences are cer¬
tainly principal causes of language change in nonindigenous languages on the periph-

The Viability of a South Asian Homeland 1 45

ery (e.g., Blust 1991). But it has also been well established that indigenous languages
tend to innovate, while emigrating ones peripheral to the homeland are often conserva¬
tive. One must conclude, then, that standard formulas cannot be applied generically:
linguistics has provided two acceptable models here that appear to conflict with each
other.

Whichever model one adopts, one is still left with the fact that the extent of conser¬
vatism or the degree of lexical and structural innovation evidenced in the various Indo-
European languages will depend on the reconstruction of the original lexicon and
morphology of the Proto-Indo-European language. Linguists disagree considerably in
their reconstructions, and few would posit an original, monolithic Proto-Indo-European
at all, since the so-called proto-language is a conventional term for a certain dialectal
continuum (D’iakonov, 1988). Accordingly, the model from which to compare later Indo-
European languages so as to determine their degree of transformation is not universally
accepted. This leaves room for debate regarding whether Sanskrit is more or less con¬
servative or archaic vis-a-vis other Indo-European languages. 7 It has already been noted
in chapter 4 that historically Sanskrit has always been considered the most conservative
Indo-European language and that some linguists would still agree with this. However,
there is a growing body of linguists, including Gamkrelidze and Ivanov themselves, who
consider some of the features long held to be archaic in Sanskrit and Greek to be later
innovations while the more simple morphology of German and Hittite more closely
represents that of the original language.

If one bears in mind the time in which he was writing, Dhar’s arguments are in¬
triguing because they challenge established assumptions and reconfigure the same data
to reverse the direction of the Indo-European movements. More significantly, he shows
considerable awareness of the linguistic theories of his day and argues in a manner just
as reasonable, logical, and coherent as that articulated by many of his Western counter¬
parts sixty years ago. 8 This intensifies the crucial issue: To what extent can the direction
of Indo-European speech movements be reversed to support the possibility of linguistic
emigrations from India as opposed to immigrations into it? Can the linguistic data ac¬
commodate the possibility that the Indo-European language might actually have origi¬
nated in India and broken up into daughter dialects that, wfth the exception of Indo-
Aryan, emigrated from the subcontinent in much the same fashion and order as linguists
would have them leave from the Volga Valley, Anatolia, Central Europe, or any other
proposed homeland? Can the position of the Indigenous Aryan school be supported by
simply reversing the generally accepted direction of linguistic flow out of the subconti¬
nent rather than into it, at least in theory?

Dialectal Subgroupings: Gamkrelidze and Ivanov’s Model

Hock (1999a) notes that “the ‘PIE-in-India’ hypothesis is not as easily refuted [as argu¬
ments claiming that Proto-Indo-European is Vedicl. ... Its cogency can be assessed only
in terms of circumstantial arguments, especially arguments based on plausibility and
simplicity” (12). There is, in other words, the issue of Occam’s razor. In Hock’s estima¬
tion (forthcoming), the consideration that Proto-Indo-European could have developed
dialectal diversity within India, “while . . . not in itself improbable . . . has consequences

146 The Quest for the Origin of Vedic Culture

which, to put it mildly, border on the improbable and certainly would violate basic
principles of simplicity” There are corollaries to such a theoretical move. According to
Hock. (1999a):

What would have to be assumed is that the various Indo-European languages moved out
of India in such a manner that they maintained their relative position to each other dur¬
ing and after the migration. However, given the bottle-neck nature of the route(s) out of
India, it would be immensely difficult to do so. Rather, one would expect either sequen¬
tial movement of different groups, with loss of dialectal alignment, or merger and amal¬
gamation of groups with loss of dialectal distinctiveness. (16-17)

Misra’s response (forthcoming) to this is that “simplicity is no argument. Migration
is not supposed to occur in a planned manner. . . . On this basis India can not be
excluded.” In other words, Occam’s razor may have no resemblance to reality. From
this perspective, it is impossible to demonstrate why the trajectories of protohistoric
languages need correspond to present-day notions of ease or simplicity, or why a sce¬
nario deemed less complicated on a theoretical diagrammatic linguistic map need repre¬
sent historical reality any better than a more complicated one. Diebold (1987), in re¬
sponse to Dyen’s principle noted earlier, articulated similar feelings: “Some languages
. . . simply will not stay put, and the migratory routes they pursue may be little affected
by such principles” (46). Occam’s razor is, however, the only tool that historians have
at their disposal when faced with a body of data, irrespective of how accurate their inter¬
pretations factually are. Hock’s point is thar Proto-Indo-European must have contained
dialectal variations prior to splitting up. He finds it very unlikely that the languages
which shared dialectal isoglosses in a hypothetical Proto-Indo-European homeland in
India would all emigrate and then resettle themselves with the same dialectal relation¬
ships they had previously, leaving only Indo-Aryan behind. It seems simpler to posit
one migration into India.

Koenraad Elst (1996) tries his hand in acnially offering a falsifiable model of how
migrations from a South Asian homeland might have occurred from within the param¬
eters of the relevant dialectal relationships:

We propose drat there is no necessary link between the fact that Sanskrit is not the oldest
form of IE and the hypothesis that India is not the oldest habitat of IE. It is perfectly
possible that a Kentum language which we now label as PIE was spoken in India, that
some of its speakers emigrated and developed Kentum languages like Germanic and
Tokharic, and that subsequently the PIE language in its Indian homeland developed and
satemized into Sanskrit. (227) 9

Elst raises the possibility of Proto-Indo-European evolving into Vedic in India itself, an
evolution which involved, over time, the loss of certain archaic Proto-Indo-European
traits. Meanwhile, other Indo-European languages left India at various stages, some of
them preserving particular Proto-Indo-European linguistic features that were not pre¬
served in Vedic (such as the maintenance of velars where Vedic had developed palatals
in Elst’s kentum-satem (nowadays denoted as centum/satsm) example, which will be dis¬
cussed later).

Let us consider the Proto-Indo-European-in-India hypothesis as a purely theoretical
linguistic exercise. I wish to stress again that this exercise becomes relevant to the field
oflndo-European smdies not as a presentation of theories claiming India to be the actual

The Viability of a South Asian Homeland 147

Indo-European homeland, but as an experiment to determine whether India can defini¬
tively be excluded as a possible homeland. If it cannot, then this further problematizes
the possibility of a homeland ever being established anywhere on linguistic grounds. If
the linguistic evidence cannot even eliminate India as a candidate, then any attempt at
archaeologically identifying the Indo-Europeans in an area supposedly preconfigured by
linguistics becomes even further complicated.

The Indo-European languages were initially divided on the basis of the velar-palatal
distribution of cognate terms into a Western branch, called the centum group (Latin
centum < PIE *kmtom ‘hundred’), 10 which had preserved the Proto-Indo-European velar
phoneme -k and a satem (Avestan satom/Sanskrit satam < PIE *kmtom ‘hundred’) branch,
which had developed a palatal phoneme -s for the same term. The centum group in¬
cluded Celtic, Greek, Italic, Germanic, and Anatolian, and the satem group, Balto-Slavic,
Indo-Iranian, Albanian and Armenian. This neat east-west division, however, was short¬
lived. A centum Indo-European language called Tocharian was found as far east as Chinese
Turkestan (Xinjiang). 11 Moreover, Melchert (1987) has argued that the Anatolian lan¬
guage Luvian is neither satam nor centum, thus questioning the heuristic value of the
entire divide. In any event, Elst’s proposal that earlier tribes could have emigrated from
India bearing the centum characteristics and, after the velars had evolved into palatals in
the Indian Urheimat, later tribes could have followed them bearing the new satem forms
(while the Indo-Aryans remained in the homeland), cannot actually be discounted as a
possibility on these particular grounds. This scenario would receive some support if the
recent report by Zoller of a centum language spoken in the North of India proves well-
founded. 12 This language, called proto-Bangani, although largely assimilated by the sur¬
rounding dialects, has supposedly preserved some centum vocabulary, especially, and
significantly, in die context of old stories.

However, there are other more significant isoglosses that need to be addressed by
an Indian or any other Urheimat, some of which cut across the centum-satem divide.
Isoglosses are bundles of common linguistic features that crisscross over language ar¬
eas. If a significant number of isoglosses coincide, the area they demarcate can be con¬
sidered a specific dialect or, if the number is sufficiently high, a language. 13 Gamkrelidze
and Ivanov (1995) have presented a temporal and spatial model for the segmentation of
Proto-Indo-European into the historically attested Indo-European dialects, which we can
apply to the Indian Urheimat to see if it fits. 14 The model (which combines the Stamm-
baum model and the wave theory model) 15 is based on isoglosses that bind, or separate,
various Indo-European languages into groups that require geographical proximity, or
segregation, at various stages of linear time. 16 Initially, in the Proto-Indo-European stage,
all the languages are united temporally and geographically, that is, they exist at the same
time and in more or less the same place, whenever and wherever they might be. 17

According to the Gamkrelidze and Ivanov model, this protolanguage initially con¬
tains two major dialect groupings, which they call A and B. 18 Group A consists of
Anatolian, Tocharian, and Italic-Celtic, group B, of Indo-Aryan, Greek, Armenian, Balto-
Slavic, and Germanic. 19 Anatolian, which is held to have uniquely preserved some very
archaic language features as 1 have discussed, is the first to break away from the home¬
land, leaving the rest of group A and group B together for a period during which they
develop some common isoglosses not visible in Anatolian. 20 Many scholars hold that
Tocharian was the next to break off (Ringe et al. 1998). After the initial departure of

148 The Quesc for the Origin of Vedic Culture

Gamkrelidze and Ivanov’s model for the segmentation of Proto-Indo-European (after
Gamkrelidze and Ivanov, 1994).

Anatolian, and Tocharian group A my parts company with group B and eventually
subdivides into the Celtic and Italic language groups that enter into protohistory.

After being separated from group A, several isoglosses in group B require that Indo-
Iranian, Greek, Germanic, and Balto-Slavic all coexisted in some degree of proximity. 21
Subsequently this group also subdivides into Balto-Slavic-Germanic and Indo-Iranian-
Greek-Armenian, 22 but in such a way that Indo-Iranian maintains a central position for
a period. This centrality allows it to share isoglosses with Slavic, on the one hand, 23
and Germanic, on the other, 24 even while remaining more closely affiliated with Greek
and Armenian. Balto-Slavic-Germanic also goes its separate way in time, and the re¬
mainder of group B, having developed some common features among its members, also
eventually breaks down into the individual Indo-lranian, Greek, and Armenian groups
that ultimately manifest in the historical record. These morphological isoglosses sepa¬
rating the various groups are further reinforced by phonemic and lexical isoglosses, which
are “unambiguous evidence for the historical reality of the dialect areas of Indo-Euro¬
pean” (Gamkrelidze and Ivanov 1995, 364). 25

While several Indo-Europeanists have found a variety of features in Ivanov and
Gamkrelidze’s account that they disagree with to a greater or lesser extent (particularly

The Viability of a South Asian Homeland

149

in their location of the homeland, the migrational routes they propose therefrom, and
their identification of specific loanwords), their basic outline of the isoglosses and group¬
ings joining the Indo-European dialects is more than adequate for the present purposes,
which is to consider to what extent dialectal geography can delimit the homeland. There
seems little in Gamkrelidze and Ivanov’s model, in and of itself, that rules out a priori,
an Indian homeland (or a variety of homelands, for that matter). Scholars such as Elst
have argued that Proto-Indo-European could have developed the dialectal distinctions
outlined earlier in South Asia itself. Anatolian could have peeled off and made its way
to its historic location from the subcontinent. The rest of group A, Celtic, Tocharian,
and Italic, could, in time, have also headed out for greener pastures. The remaining
dialects could have continued to evolve their defining characteristics in India. The Indo-
Iranian tribes, or speakers, could have been neighbors with the proto-Greek and Arme¬
nian speakers, on one side, and the proto-Balto-Slavic and Germanic speakers, on the
other, resulting in whatever shared isoglosses need to be accounted for with these vari¬
ous groups. Satemization, for example, could have been an isogloss that partially spread
throughout this area but without reaching Germanic on one side, or Greek on the other.

In time, tire Balto-Slavic-Germanic speakers moved on, in turn, leaving the remaining
three dialects—Greek, Armenian, and Indo-Iranian—to continue to interact linguistically,
explaining the close similarities between them. Eventually, even Greek and Armenian
departed for the Northwest, leaving Indo-lranian behind. The Iranians had the least dis¬
tance and were the last to move: any hypothetical eastern homeland could, perhaps, have
included parts of northeastern Iran. 26 The out-of-India model as outlined by Elst is really
just a co-option of Gamkrelidze and Ivanov s model but with an even more southeasterly
point of origin. This scenario basically differs from the more common Caspian Sea home¬
land model by postulating a linear and sequential initial point of origin rather than one
radiating out in a more circular fashion. Ultimately, there is no internal mechanism in the
direction of language spread that can a priori discount such a model, even if it disregards
Dyen’s principle that language family homelands are more likely to be located in the area
where there is maximum contiguity between the cognate languages of the family (or, to
put it differently, the area from where the minimum number of migrations are required to
connect the distributions of drese cognate languages).

Or course, many other issues remain to be dealt with. Migration routes need to be
postulated whereby loanwords into the Indo-European languages from other language
families can be accounted for; Gamkrelidze and Ivanov’s whole case is heavily depen¬
dent on Semitic and Caucasian loans which they have identified in Indo-European. Their
etymologizing and interpretations of the implications of these loans are, in turn, con¬
tested by their detractors who propose different homelands. Ultimately, since it is clear
that our study group is interested only in the indigenousness of the Indo-Aryans, and
not in the saga of the other Indo-European speakers, Indigenous Aryanists are not likely
to contest any migration details proposed for the other languages unless they militate
against a South Asian homeland. Thus, Elst is quite happy to note that, in Gamkrelidze
and Ivanov’s model (1996),

all the migrations required are the same as in the IUT [Indian Urheimat theory], which

would equally bring the IE tribes to the East of the Caspian sea first, then let them move

on from there. . . . From an IUT viewpoint, all this falls neatly into place. . . . [Italic,

150 The Quest for the Origin of Vedic Culture

Celtic, Germanic, Baltic, and Slavic] entered Europe through Ukraine, north of the Black

Sea, while Greek and Palaeo-Balkanic parted company with Armenian on the South coast

of the Caspian Sea. (237-239)

In fact, Gamkrelidze and Ivanov’s migration routes have received criticism from schol¬
ars on linguistic grounds (e.g., D’iakonov 1985). Every homeland theorist runs into
problems with Occam’s razor at some point or another. But the point holds good: die
phonemic, morphological, and lexical relationships between the various Indo-European
languages, in and of themselves, are not sufficient as data to pinpoint the original home¬
land; the diachronic and synchronic relationship between the various dialects can be
accounted for in a variety of ways and from a variety of homelands.

Ultimately, apart from non-IE influences, such as loanwords, there is nothing in the
languages themselves that can give any indication of how far the languages might have
traveled to reach their historic destination. In the words of Latham (1862), who, in the
mid-nineteenth century was attempting to challenge the then very entrenched idea that
Asia was the homeland of the Aryans (the exact reverse of what the Indigenous Aryan
school is doing today): “A mile is a mile and a league is a league from which ever end
it is measured, and it is no further from the Danube to the Indus than it is from the
Indus to the Danube” (612). Of course, it can be pointed out that Elst’s model appears
clumsy, since “it is easier to accept that remotely related languages developed separately
within the same region rather than that several unconnected waves of migration to that
same region brought remotely related languages with them” (Pejros 1997, 152). Be that
as it may, none of these arguments can be used convincingly to insist on an external
origin for the Indo-Aryan languages.

In my opinion, the most serious objection against a South Asian homeland from
the perspective of the evidence being considered in this chapter is not the dialectal align¬
ment of the languages but the homogeneity of Indo-Aryan in the subcontinent. If Proto-
Indo-European had developed dialectal isoglosses in India in the manner outlined by
Elst why would all the different dialects have emigrated to eventually become distinct
languages, leaving only one solitary language behind? Why did some of the dialectal
variants germinating in our hypothetical Proto-Indo-European in South Asian not re¬
main to develop into other non-Indo-Aryan, Indo-European languages on the subcon¬
tinent itself? After all, Sanskrit developed into a variety of mutually incomprehensible,
distinct languages in the historic period, such as Braj, Bengali, Punjabi, Gujarati, Marathi,
and so on. So why would PIE not likewise have developed into other distinct Indo-
European languages in addition to Sanskrit in the subcontinent itself in the protohistoric
period? Why did they all emigrate?

Of course, one could argue that there may have been other ancient South Asian Indo-
European languages that suffered ‘language death’ because Old Indo-Aryan (Vedic)
became a culturally elite language that subsumed them, but unless proto-Bangani be¬
comes accepted as a genuine non-Indo-Aryan, Indo-European language, there would
need to be compelling evidence to consider such pleading (although Old Indo-Aryan
did contain dialectal variants; see, for example, Emeneau 1966). Another argument is
that there may have been less language variety on the western side of the Indo-Euro¬
pean family in the second millennium b.c.e. at a time contemporaneous with Vedic
and that therefore Indo-Aryan might have been less conspicuously homogeneous in the

The Viability of a South Asian Homeland

151

Southeast. Part of this western heterogeneity, after all, is established on the basis of
languages most of which are first attested well after Vedic at a time when the Indo-
European South Asian languages had also developed some degree of heterogeneity; we
have no texts or sources from Europe informing us of the degree of differentiation of
the European Indo-European languages in 1500 b.c.e. Our oldest actual empirical evi¬
dence for the major families in Gamkrelidze and Ivanov’s group A is the Italic group in
the form of inscriptions that may date to the seventh cenmry b.c.e. well after the most
conservative date assigned to the Rgveda. The Germanic group is not attested until the
Runic inscriptions of the third century c.e.; Insular Celtic (in the form of Irish Ogam),
in the fourth century c.e. (although there are Continental Celtic languages which are
earlier, as well as some Gaulish names and inscriptions from prior to the common era;
Slavic (in the form of Old Church Slavonic) from the ninth century c.e.; and Baltic (in
the form of Old Prussian) from the fourteenth century c.e. In 1200 b.c.e. we only have
solid evidence for the existence of Greek, Anatolian, Iranian, and Indie. However, de¬
spite the comparative lateness of evidence for most of the European languages, there are
reconstructable linguistic grounds (e.g., Ringe ct al., 1998) to suggest that Anatolian
was the first to separate from tire Indo-European collective, followed by Tocharian and
then Italo-Celtic, which does indicate differentiation in the family before Indo-Iranian
took its distinctive form.

Nonetheless, it is relevant to note, in this regard, that although the Veda and older
sections of the Avesta are so close linguistically, had these old texts not been preserved
for posterity and linguists had only the modern Iranian and Indo-Aryan languages as
data with which to plot language relationships, there is no certainty that a grouping of
these modern languages into an Indo-Iranian subfamily would have been evident. Dyen
et al.’s (1992) lexicostatistical analyses, which are based on an analysis of modern lan¬
guages only, show a failure to distinguish an Indo-Iranian group; the New Indo-Aryan
languages in particular, demonstrate the same affinity with the European languages (Italic,
Celtic, Germanic, and, especially Balto-Slavic) as with the Iranian ones. In other words,
in the modern situation, there is no obvious homogeneity among the eastern languages—
we just happen to have Vedic and Avestan texts informing us that there was such ho¬
mogeneity in the ancient period.

Nichols’s Sogdiana Model

In any event, a less hypothetical explanation for this eastern homogeneity, and one
that may partly address Hock’s concern, noted above, has recently surfaced. Johanna
Nichols presents an alternative model for the epicenter of the Indo-European linguis¬
tic spread that could possibly be reformatted to accommodate a South Asian home¬
land (with perhaps less effort than that involved in the co-option of Gamkrelidze and
Ivanov’s model and more in accordance with Dyen’s principle). The model also ad¬
dresses the issues raised earlier concerning center and periphery. Nichols (1997) ar¬
gues for a homeland “well to the east of the Caspian Sea . . . somewhere in the vicin¬
ity of ancient Bactria-Sogdiana” (123-1 37). Since this area overlaps the territory being
considered here for the purpose of argument, namely, a Northwest Indian/Pakistan/
Afghanistan homeland, it merits particular attention. Nichols’s theory is partly predi-

152

The Quest for the Origin of Vedic Culture

cated on loanwords between Indo-European and other language families that were
discussed in the previous chapter.

Nichols holds that the principle that the area of greatest genetic diversity of a lan¬
guage family is indicative of its locus of origin is demonstrably false for the languages of
central Asia. She cites Iranian, which spread over enormous stretches of Asia in ancient
times, and Turkic, which likewise spread over major portions of Asia, as examples of
languages whose greatest diversity occurred in refuge areas on the western periphery of
their point of origin. Nichols then draws attention to Dyen’s definitions of a homeland
as a continuous area, a migration as a linguistic movement out of this area that causes
it to become noncontinuous (in other words, which separates itself from this area), and
an expansion as a movement that enlarges this area. As noted previously, Dyen’s home¬
land-locating formula discourages migration and prioritizes contiguity.

In Nichols’s Bactrian homeland, Proto-Indo-European expands out of its locus, even¬
tually forming two basic trajectories, appearing, on a language map, like two amoebic
protuberances bulging out from a protoplasmic origin. The language range initially ra¬
diates westward, engulfing the whole area around the Aral Sea from the northern steppe
to the Iranian plateau. Upon reaching the Caspian, one trajectory expands around the
sea to the north and over the steppes of central Asia to the Black Sea, while the other
flows around the southern perimeter and into Anatolia. 27 Here we have a model of a
continuous distribution of Proto-Indo-European—which has been defined as being, in
reality, a dialectal continuum—covering a massive range from where the later historic
languages can emerge, without postulating any migrations whatsoever. By the third or
second millennium b.c.e. we have the protoforms of Italic, Celtic, and perhaps Ger¬
manic in the environs of central Europe (and presumably Balto-Slavic as well), and the
protoforms of Greek, Illyrian, Anatolian, and Armenian stretching from northwest

The Viability of a South Asian Homeland

153

Mesopotamia to the southern Balkans (Nichols 1997,134). Proto-Indo-Aryan was spread¬
ing into the subcontinent proper, while proto-Tocharian remained close to the original
homeland in the Northeast.

As this expansion was progressing into Europe, a new later wave of Indo-European
language, Iranian, was moving behind the first language spread (Nichols does not indi¬
cate an exact time frame except to note that this spread was posterior to the first expan¬
sion, presumably some time in the second millennium b.c.e .). 28 Nichols also does not
indicate the exact point of origin of Iranian, but one might assume that it was the evo-
lute of Proto-Indo-European that emerged from more or less the same locus, since it
follows the same trajectories taken by the preceding waves of Indo-European. 29 How¬
ever, sweeping across the steppes of central Asia, the Caucasus, and the deserts of north
Iran, the Iranian dialects separated tire two preceding trajectories—which until that time
had formed a continuum—into two noncontiguous areas (one in central Europe to the
north of the Caspian Sea, the other in Anatolia to its south). Along the same lines,
Iranian also separates Tocharian from the other languages (which would eventually
become completely severed from other Indo-European languages by the incursion of
Turkic into this area). In time, the two original trajectories coincided in the Balkans.
The southern trajectory had formed a continuous chain of Dacian, Thracian, Illyrian,
Greek, and Phrygian spreading from West Anatolia to the Danube plain (Nichols 1997,
136). 30 From the northern trajectory, Italic spread to Italy from central Europe, and
Celtic to its historic destination, followed in time by Germanic, which was followed in
turn by Balto-Slavic. All of these languages spread by expansion—there are no migra¬
tions throughout this whole immense chronological and geographic sequence. 31

The corollary of Nichols’s model is that it portrays the homogeneity of Indo-lranian
and the heterogeneity of the western languages in a new light. The assumed variegation
of the western languages is only due to the fact that the later Iranian language had spread
and severed the contiguity of the northern and southern Indo-European trajectories (which
had previously formed an unbroken continuity around the east coast of the Caspian),
thereby making them appear noncontinuous while leaving behind Indo-lranian and a
stranded Tocharian to the east. The variegation of western languages is actually due to
their situation on the western periphery of the original locus, or homeland. This model
might also address the issue of why Proto-Indo-European did not evolve into more dia¬
lects in the putative homeland: the later westward spread of Iranian obliterated all of
the eastern parts of the protocontinuum except for Indo-Aryan to its east and the iso¬
lated Tocharian to the northeast. 32

From the perspective of dialect geography, the question arises whether Nichols’s home¬
land model, if enlarged somewhat, or relocated a little toward the southeast, could be
applied to the hypothetical Bactrian/Pakistan/Northwest Indian homeland. From this
perspective, one would have to postulate a scenario along lines similar to the following:
Proto-Indo-European could have evolved in west South Asia into a continuum of dia¬
lects that radiated westward in the protohistoric period, covering an unbroken area from
the desert of Iran to the steppes of central Asia. In this continuum, Indo-lranian held
a central position. More specifically, the evolving Proto-Indo-Aryan elements in Indo-
lranian held a central position, with the evolving Proto-Iranian dialectal elements mani¬
festing on the western perimeters of the Indo-Aryan core as the part of the continuum
nearest the locus of origin. 33

154 The Quest for the Origin of Vedic Culture

This western periphery of Protolndolranian was flanked on its northern side by
Balto-Slavic, Germanic, Celtic, Italic, Tokharian, and Anatolian. Anatolian may have
peeled off earliest to resurface in its destination in the Near East, while Tocharian ex¬
panded farther to the northeast and off the western path of subsequent language spread,
becoming a language isolate (or separate, in Dyen’s terms). 34 Italic and Celtic continued
to expand as the vanguard of the northwest trajectory over the steppes. They formed a
continuum and shared an isogloss with Germanic— tt > tst > ts > ss (Hock forthcom¬
ing)—which, in turn, was followed by, and shared isoglosses with, Balto-Slavic. 35 Balto-
Slavic was immediately contiguous with the northwestern part of the Indo-Iranian dia¬
lects (which, as noted, were central to this whole Indo-European continuum), producing
a further isogloss between these particular dialects (a merger of velars and labiovelars).

On the southwestern side of tire core area, Armenian was immediately contiguous
to Indo-Iranian and was in turn connected on its western periphery with Greek. 36
Satemization was an isogloss drat spread through a core area of Indo-Iranian and its
adjacent dialects on both sides—to the northwest and the southwest. On its southwest¬
erly side, it affected Armenian and, in diminishing degrees, the transition areas of
Thracian, Pelasgian, and Phrygian 37 but without reaching Greek, which was farther south¬
west. The change also rippled through a transition area of Balto-Slavic, to the northwest
of Indo-Iranian, but without reaching Germanic, which was farther northwest again. 38
Another isogloss, that of the preterit augment, was shared by Indo-Iranian, Armenian,
and Greek. 39 Dhar’s comments about Indo-Aryan conservatism and the influence of
substratum on the languages of the western expansion of Indo-European are not in¬
compatible with this model. It seems that other significant isoglosses can also be ac¬
counted for in this system. 40

This model seems less clumsy than Elst’s co-option of the Gamkrelidze and Ivanov
model, which involves a kind of staggered exodus of several distinct groups and is sub¬
ject to Hock’s criticism noted earlier. In D’iakonov’s view (1990, 156), the Indo-Euro¬
pean language speakers never left their homeland, wherever it may have lain; rather,
there was the constant spread of an increasing population to neighboring, peripheral
territories. I will leave it to linguists more qualified than myself to fine-tune the details
of Nichols’s model or point out its irreconcilable flaws, which can then be compared
with the linguistic idiosyncracies that have been pointed out in all the other homeland
models. 41 And, of course, there are Indo-Iranian loans in Finno-Ugric as well as in the
Caucasian languages that need to be accounted for. Nichols does not directly address
the issue of Indo-Iranian loans into Finno-Ugric, but one might assume that her model
allows for these loans to have occurred when the Iranian languages swept over the steppes
on the heels of the first wave of Indo-European languages.

Conclusion

The immediate objections to Dhar’s, Elst’s, and Nichols’s linguistic theories are likely
to come from archaeologists. Indeed, even prior to reading the formal publication of
Nichols’s arguments, Mallory was moved to outline an archaeological response to the
very idea of an eastern homeland. He claims that there is little evidence of urban-steppe
interaction, which Nichols’s model would require, before about 2000 b.c.e., which is

The Viability of a South Asian Homeland 155

far too late for Indo-European dispersals. No obvious candidate in the archaeological
record has been found that might correspond to a movement of peoples from Bactria
(or the Northwest of the subcontinent) to the west in an appropriate time frame. Such
objections, of course, are based on the assumption that language spread can be traced
in the archaeological record. There has been a recent chorus of objections to this idea.
Mallory himself notes in an earlier publication that “as the IE homeland problem in¬
volves a spatial definition of a prehistoric linguistic construct, the utility of any other
discipline, such as archaeology, depends on whether a linguistic entity can be translated
into something discernable in the archaeological record. In short, any solution not purely
linguistic must involve some form of indirect inference whose own premises are usu¬
ally, if not invariably, far from demonstrated” (Mallory 1997, 94).

Indeed, as will be discussed later, scholars have consistently pointed out that there
are no archaeological traces of the movements and migrations of a number of ethnic
groups, such as the Huns, or of the Helvetii, and one would never have known that
such occurrences had ever taken place had they not been documented in historical records.
Mallory (1989, 166) has also noted that there is no archaeological evidence that can be
correlated with the introduction of Gaelic into Scotland; likewise with the movements
of the Slavs into Greece and the Galatians into central Anatolia (Crossland 1992, 251).

The argument that there is no archaeological evidence to substantiate any hypotheti¬
cal linguistic spread from the east to the west is especially unlikely to find much favor
on the subcontinent. As we shall see in chapters 10 and 11, Indian archaeologists have
long pointed out that there is no consistent archaeological culture that can be traced
across central Asia to penetrate into the subcontinent that might be correlateable to the
Indo-Aryans; they have long been informed by linguists that there need not be any ar¬
chaeological evidence of such movements, since language can spread without any change
in the material culture. They are hardly likely to now accept that such archaeological
invisibility applies only to immigration into the subcontinent but not to emigration out
of it. Actually, no South Asian archaeologists that I know of have ever attempted to
argue for an emigration—as I noted earlier, it is the details of the Indo-Aryans that are
of concern, not those of the other language family members. While this may be some¬
what myopic and unsatisfactory from an Indo-Europeanist’s point of view, Indigenous
Aryanists have a right to expect that, at least in theory, whatever arguments are brought
forward by Western scholars as holding good for immigration in the archaeological record
should equally be expected to hold good for any emigration, however hypothetical. Let
us not forget that Indian scholars have every reason to suspect a history of Western bias
in the handling of the Indo-European problem. I will be turning to the archaeological
record in the next chapters.

The point of all this is not to argue that Dhar’s arguments, or Elsts’s co-option of
Gamkrelidze and Ivanov, or Nichols’s enlarged model actually support or suggest a South
Asian homeland. The point is merely to suggest that these types of linguistic data are
unequipped to exclude such possibilities. The point is also to provide some sense of
how those suspicious of the entire homeland-locating enterprise can approach the ma¬
terial with different presuppositions and perspectives, rearrange the same data, and
assemble entirely different hypothetical version of proto-historic events. At this point,
scholars will have to resort to Occam’s razor to judge between the various homeland
claims—which scenario requires the least amount of special pleading in accounting for

156 The Quest for the Origin of Vedic Culture

the whole range of data. As the two-centuries-old history of the Indo-European problem
demonstrates, however, there is not likely to be more consensus in this regard among
scholars in the present dian there has been in the past. Moreover, the judgment regard¬
ing which case requires more special pleading is likely to be viewed very differently by
many Indian scholars than by some Western ones.

The basic thrust of the previous chapters has been to show some of the problems
that have arisen in the attempt to pinpoint the origins of die Indo-European—and of
their offshoots, the Indo-Aryans—through linguistic methods. At this point I should
note that relatively few Indo-Europeanists nowadays have an active interest in the mat¬
ter of the urheimat. For decades, scholars have realized that the difficulties with the lin¬
guistic evidence are considerable enough to make each and every conclusion based on
it problematic; however, the issue remains of considerable concern to many scholars of
South Asia. In the next three chapters, we will turn our attention to the archaeological
evidence, bearing in mind that the Indo-Aryans are a linguistic and philological entity.
Archaelolgists can only attempt to trace speakers of a language family when linguistics
and philology have provided them with clear and unambiguous information about an
identifiable material culture that can be associated with such speakers.

The Indus Valley Civilization

In the words of one of India’s leading nationalist historians: “There is one curious fact
in regard to the beginnings of Indian history. For the Indus Valley culture, we have
abundant archaeological data, but no written evidence. For the early Vedic culture we
have abundant written evidence but no archaeological data” (Majumdar 1959, 6). The
Indus Valley Civilization covers about a million square miles, yet there is no consensus
regarding who its inhabitants were. The archaeological sites of the Indus Valley occupy
much of the same geographic horizons known to the composers of the Vedic hymns. It
seems understandable that many scholars might be tempted to fit the two together.
Numerous books and articles have attempted to fit Vedic descriptions of culture, soci¬
ety, and religion into the ruins of the Indus Valley. Such endeavors, for the most part,
take great interpretative liberties, and I will only touch upon a few of them here to give
a sense of some of the issues involved, since some general comments on the Indus Val¬
ley are unavoidable, in any discussion of Indo-Aryan origins. It is essential for the In¬
digenous Aryan case that the Indus Valley Civilization be an Indo-Aryan one for obvi¬
ous reasons. The issue that needs to be addressed in this chapter, accordingly, is: Did
the Indus civilization precede the Vedic one, did it follow the Vedic one, or were the
two contemporaneous?

Indra Stands Accused

Up until the discovery of the Indus Valley Civilization in 1922, images of virile, blond,
northern tribes swooping across die mountain passes on chariots and overpowering the
primitive and ill-equipped natives they found on their way were presented as the stan-

157

158 The Quest for the Origins of Vedic Culture

Area of the Indus Valley Civilization.

dard version of the early history of the subcontinent. The 1920 edition of The Oxford
Student’s History of India (reprinted 1933), for example, states that “as they advanced the
Aryans [who were ‘tall, fair, long-nosed and handsome’] subdued, more or less com¬
pletely, the ‘aborigines’ [‘short, dark, snub-nosed and ugly’], whom they called Dasyus,
and by other names” (Smith 1933, 25-26). Likewise, the Vedic India volume of the
popular Story of the Nations series informs its readers that “the natives . . . belonged to
a black, or at least a very dark race, and everything about them, from their color and flat
noses, to their barbarous customs, such as eating raw or barely cooked meat, and their
Shamanistic goblin-worship, was intensely repulsive to the handsome, gentler mannered
and, to a certain degree, religiously refined and lofty-minded Aryas” (Ragozin 1895, 113).

Then, in 1922, the Indus Valley Civilization was discovered. 1 Sir John Marshall
(1931) describes the civilization of these pre-Aryan natives of India in his official ac¬
count of the archaeological excavations at Mohenjo-Daro carried out between 1922 and

1927:

The Indus Valley Civilization 159

Hitherto it has commonly been supposed that the pre-Aryan peoples of India were . . .
black skinned, flat nosed barbarians. . . . Never for a moment was it imagined that five
thousand years ago, before the Aryans were heard of, Panjab and Sind . . . were enjoying
an advanced and singularly uniform civilization of their own . . . even superior to that of
contemporary Mesopotamia and Egypt. . . . there is nothing that we know of in prehis¬
toric Egypt or Mesopotamia or anywhere else in western Asia to compare with the well-
built baths and commodious houses of the citizens of Mohenjo-dara. . . . nothing that we
know of in other countries at this period bears any resemblance, in point of style, to the
miniature faience models . . . which . . . are distinguished by a breadth of treatment and
a feeling for line and plastic form that has rarely been surpassed in glyptic art. (v-vii)

Edmund Leach (1990) comments wryly on the academic reaction to Marshall’s disclo¬
sures:

Common sense might suggest that here was a striking example of a refutable hypothesis
that had in fact been refuted. Indo-European scholars should have scrapped all their
historical reconstructions and started again from scratch. But that is not what happened.
Vested interests and academic posts were involved. Almost without exception the schol¬
ars in question managed to persuade themselves that despite appearances, the theories of
the philologists and the hard evidence of archaeology could be made to fit together. The
trick was to think of the horse-riding Aryans as conquerors of the cities of the Indus
civilization in the same way that the Spanish conquistadores were conquerors of the cities
of Mexico and Pent. . . . The lowly Dasa of the Rig Veda, who had previously been thought
of as primitive savages, were now reconstructed as members of a high civilization. (237)

Scholars like Stuart Piggott and Sir Mortimer Wheeler are especially targeted as the
quintessential creators of such images of incoming Aryan aggressors destroying this newly
found civilization of the erstwhile lowly Dasa: “Tangible archaeological evidence of the
Aryan conquest of India consists of nothing but the ruins of the cities they wrecked”
(Piggott 1952, 285). Pivotal to such theories was the discovery of thirty-seven skeletons
in various locations of Mohenjo-Daro—especially of a huddled group of half a dozen
skeletons sprawled in one of the lanes in the city. Two of these skeletons had marks on
their skulls suggestive of a blow from a sharp object. On the basis of these skeletons,
Wheeler (1968) confidently stated that “the end of Mohenjo-daro . . . was marked by a
massacre as [this] evidence quite unquestionably indicates” (83). 2 This evidence was
then juxtaposed with so-called citadels found in several sites such as Mohenjo-Daro,
Harappa, and Kalibangan, which Wheeler took to be fortified mounds—the pur of the
Rgveda (78). With Indra, whose epithet in the Rgveda is purandara ‘fort-destroyer’, as
his chief protagonist, Wheeler had a dramatic script that he could have marketed in
Hollywood. “Indra stands accused” was his lighthearted, but later regretted, caricature
of the principal culprit behind the demise of the great civilization (Wheeler 1953, 92). 3

Scholars soon began to react against Wheeler’s version of events. In time, most scholars
judged that “Indra stands completely exonerated” (Dales 1964, 42; See also Srivastava
1984, 441). George Dales (1964) pointed out the obvious: “Where are the burned for¬
tresses, the arrowheads, weapons, pieces of armor, the smashed bodies of the invaders
and defenders? Despite the extensive excavations at the largest Harappan sites, there is
not a single bit of evidence that can be brought forth as unconditional proof of an armed
conquest and destruction on the scale of the Aryan invasion” (38). Not a single one of
the thirty-seven skeletons was found within the area of the so-called citadel, which pre-

160 The Quest for the Origins of Vedic Culture

sumably would have been the locus of die heaviest fighting in the siege of a city. Besides
this, the celebrated group of skeletons were found to belong to a period posterior to the
abandonment of the latest stage of the city (38). Moreover, Kenneth Kennedy (1994,
248), who inspected thirty-four of the skeletons, found only one revealed a cranial le¬
sion that might have been inflicted by a weapon; the marks on the remaining skulls,
apart from one that had a healed wound mark unconnected with the cause of death,
were cracks and warps caused by erosion, not violent aggression. Kenoyer (1991b) sums
up the situation: “Any military conquest that would have been effective over such a
large area should have left some clear evidence in the archaeological record. . . . evi¬
dence for periods of sustained conflict and coercive militaristic hegemony is not found”

(57).

Few archaeologists today refer to Aryan aggression in connection with the demise of
die Indus Valley, although occasionally the old paradigm stirs again; as recently as in
Possehl’s 1993 edition of Harappan Civilization F. R. Allchin (1993) still ponders, al¬
though more hesitantly than in 1968: “Would the attackers have been Aryan! . . . there
is no inherent impossibility in such a thing” (389). For the most part, however, various
alternative ecological or socioeconomic theories have been accepted regarding the gradual
abandonment of the vast civilization. 4 Few scholars are ready to attempt a retrial of Indra
and his Aryans. So what were our invading, or intruding, Aryans doing if they were not
destroying the cities of the Indus Valley? A growing number of Indian archaeologists
believe that the Indus Valley could have been an Indo-Aryan civilization or, at least,
that the two cultures could have coexisted. A variety of evidence has been brought for¬
ward to support this possibility, some of which will be reviewed in the following pages.

The Religion of the Indus Valley

As a result of excavations at the site of Kalibangan on the banks of the dry Ghaggar
River in North Rajasthan, a number of Indian archaeologists, including Migrationists
such as as Parpola and Allchin, accepted that the findings there “are highly suggestive
of an Indo-Iranian, if not more specifically Indo-Aryan, element in the culture of the
period covered” (Allchin 1993, 388; see also Parpola 1988 149). Allchin proposes that
intrusive Indo-Aryan groups had synthesized with indigenous Indus dwellers, resulting
in these possible Indo-Aryan cultural traces that had been uncovered in this Indus site.

These findings, which have been termed “fire altars” or “ritual hearths,” were found
in both public and residential locations. In a nonresidential area of Kalibangan, a series
of raised platforms was excavated, each of which was accessible by a flight of stairs.
Atop one of these, a row of seven clay-lined pits was discovered, each one measuring
seventy-five by fifty-five centimeters and containing traces of ash, charcoal, and the re¬
mains of a clay stele. The layout of the pits was such that the officiator would have been
obligated to sit facing east (Lai 1984, 57). According to die excavators, this parallels the
seven dhisnya hearths of the Vedic soma sacrifice where the priests sit to the west of the
hearths facing east. Most significantly, a short distance away from these altars in
Kalibangan were found a well and the remains of a few bath pavements on the same
platform, which “clearly suggest that ceremonial bathing constituted a part of the ritual”
(57). On another nearby platform were the remains not only of a fire altar and well but

The Indus Valley Civilization 161

also of a rectangular brick-lined pit containing bovine remains and antlers “evidently
representing some kind of sacrifice” (57). The gates leading to the whole area where the
platforms were situated were flanked by salients, and access could only be attained via
steps, thus precluding the entrance of vehicular traffic. All this “may perhaps . . . appro¬
priately be termed a ‘Temple-complex’” (58).

In another part of the town was found another structure built of mud bricks. On top
of this was an impressive wall that enclosed a room containing four or five more of the
same types of fire altars. No other building existed on this mound, nor was any of the
usual occupational debris found, suggesting that “the lonely structure with the altars
was used for ritual purposes” (Thapar 1975, 28). Domestic fire altars were also found
in numerous residential houses in the “Lower Town.” In many of these houses a room
seemed to have been set aside especially for the fire altar, which was renewed repeatedly
as the working level went up (Lai 1984, 58). The chronology of this period was about
2300 to 1750 b.c.e. The official report of this excavation has yet to be published, and,
unfortunately, Thapar has since passed away. I should note that Thapar did believe in
Aryan migrations (pesonal communication).

The excavator of Lothal, S. R. Rao, found as many as six fire altars similar to the
ones in Kalibangan in different blocks of the Lower Town. From their layout and con¬
struction, he denies that they could have been domestic ovens and that “it is obvious
that they could not serve any other purpose than a ritualistic one” (Rao 1993, 175). s In
one of these, a charred bovine mandible was found. Nearby another much larger altar,
which Rao (1979) suggests was used for “community fire-worship,” a terra-cotta “ladle”
was found that “must have been used for pouring clarified butter into the sacrificial
fire” (176). Rao precludes possible suggestions that the Indo-Aryan presence was in the
final stages of occupation of the town by stating that “it is not only in the final stages of
Kalibangan but also in the early stages of Harappa culture at Lothal and Kalibangan
that altars for fire-worship and animal sacrifice were built and made use of” (173). In
addition to Kalibangan and Lothal, H. D. Sankalia (1974, 350) refers to a fire altar
noticed by Casal at Amri, and S. R. Rao (1993, 173) refers to similar reports from Bisht
at the Banawali site. Allchin and Allchin (1982) are prepared to accept that, what they
prefer to call “ritual hearths,” occur in the beginning of the Harappan period, suggest¬
ing an Indo-Aryan presence in the “still flourishing Indus civilization” (303).

Not everyone accepts Rao’s and Thapar’s identifications. Dhavalikar (1995) won¬
ders whether we are dealing with fire altars or with fire pits that could have been used
for cooking or baking. He finds them very similar in size, plan, and shape to his exca¬
vations at Inamgaon in the Deccan. Indeed, he sees similarities between them and cooking
pits used in Maharashtrian villages to this day. For Dhavalikar, the clay stele in the
center of the pits, noted by the excavators, “bears a striking resemblance with the clay
tavd . . . that is in use in Maharasthra . . . which was obviously for baking bread” (7).
In this view, “since the Kalibangan ‘fire altars’ are identical in every respect with those
in Inamgaon, their association with the religious beliefs of the people becomes doubt¬
ful” (7). One would also have to note that Lai’s identification of these altars as Vedic
seems to be primarily influenced by the fact that there were seven of them, thereby
paralleling the number of hearths in various Vedic sacrifices. However, while this is
correct, these sacrifices do not just consist of these seven hearths but include a variety of
other hearths as well, none of which were unearthed in Kalibangan.

162

The Quest for the Origins of Vedic Culture

Much has been written on the religion of the Indus people; the implications of in¬
terpreting Indus artifacts as Vedic or non-Vedic are obvious. Rather than burden the
reader with an overly detailed exposition of views on this matter, a brief glance at some
of the creative interpretations inspired by the so-called Pasupati proto-Siva seal will give
some sense of the series of assumptions that are often made in assigning meaning to
innate archaeological objects. The seal consists of what appears to be an ithyphallic fig¬
ure on some kind of a seat in yogic posture with arms resting on knees and crowned
with a horned headdress. The figure is surrounded by a number of animals, and there
is an inscription above the figures.

Sir John Marshall (1931) was the first to volunteer an identification. Taking the fig¬
ure to be “recognizable at once as a prototype of the historic Siva,” he assigned the
name—Pasupati, ‘king of the beasts’—to the seal in view of the animals surrounding the
figure. The animals are a tiger, elephant, rhinoceros, buffalo, and two deer. He believed
that the figure had three faces, which is a feature sometimes ascribed to certain forms of
Siva. Moreover the yogic posture suggested to Marshall the ascetic nature of this deity,
while the horned headdress conjured up associations with Siva’s trident. Marshall was
convinced that the Indus Civilization existed prior to the entry of the Indo-Aryans, and
that Siva was a pre-Aryan, Dravidian god who was later co-opted into the Vedic pan¬
theon (54). Marshall’s identification was to influence all subsequent interpretations.

The Indus ‘Proto-Siva’ seal.

The Indus Valley Civilization 163

Other scholars who also posit a Dravidian affiliation for the Indus Valley have inter¬
preted the seal differently. Hiltebeital (1978) developed a case for considering the horned
figure to be Mahisa, the buffalo demon toe of the goddess. He located the goddess herself
as represented in the form of a tigress—one of the animals surrounding the figure—an
animal frequently depicted as the mount of the goddess. Sullivan (1 964) also favored a
goddess identification, albeit on different grounds, and argued that the so-called erect
phallus was really a girdle such as is found only on female figurines. Fairservis (1992)
was prepared to go further and tried his hand at a Dravidian translation of the inscrip¬
tion. He held that the seal referred to “a paramount chief named Anil ... a primary
chief of the four sodalities, each one represented by one of the animals” (200). Parpola
(1994) feels that the “so-called ‘yoga’ posture may simply be an imitation of the Proto-
Elamite way of representing seated bulls” (250). His Dravidian reading of the seal is
“min-a dl ‘the man (or servant) of (the god represented by) rnin,’” which he considers in
this case to mean fish (188). He finds the animal representations in the seal best re¬
semble those associated with Varuna, an entity associated with the aquatic themes that
he finds prominent in the Indus religion.

A number of other scholars have also taken this seal as a Vedic deity, but, unlike the
authors, mentioned earlier, who understand Siva, Varuna, and the goddess as being
pre-Aryan deities who were later amalgamated into the Vedic pantheon, this group believes
that their Vedic identifications point to the Indo-Aryan identity of the Indus Valley
Civilization. S. R. Rao (1991) reads the inscription as ra-ma-trida-osa, “conveying the
sense of being ‘pleasant and shining (or burning) in three ways’. The three-headed deity
who is burning or shining in three ways is none other than Agni conceived in his three
forms” (288). The animals, for him, represent different clans that have accepted the
supremacy of the fire god. M. V. N. Krishna Rao (1982), in contrast, believes Indra is
represented by the figure. In terms of method, he finds reason to bypass the tiger, on
the grounds that it is somewhat larger and more prominent than the other animals, as
well as the two deer, on the grounds that they are seated below the main figure. He then
takes the first phoneme of the Sanskrit terms for the remaining three animals and the
first phoneme of the word ‘man’ nara (which he somehow feels deserves to be repre¬
sented twice) and construes the term makhanasana, an epithet of Indra.

S. P. Singh (1988-89) also identifies the central figure as Rudra, who is the protoform
of Siva in the Rgveda. However, his method is based on identifying the animals sur¬
rounding the figure with the Maruts, who are referred to as the sons of Rudra in hymn
1.64. His grounds for this, in turn, are that this hymn describes the Maruts as bulls of
heaven, eating up the forests like elephants, roaring like lions, beauteous as antelope,
and angry as serpents. E. Richter-Ushanas (1997) considers the central figure to be the
sage Rsyasrhga, ‘the sage with horns’, who officiated over the sacrifice of King Dasaratha
in the Ramayana. He connects the animals with the four seasons and finds similar motifs
on the Gundestrup cauldron discovered in Denmark (Taylor 1992; as a side note, Talageri
[1993] and Rajaram and Frawley [1995] see this cauldron as persuasive evidence point¬
ing to India as the original home of the Indo-Europeans). Feuerstein, Kak, and Frawley
are quite happy accepting Marshall’s identification of the seal except with an under¬
standing that Siva is an Indo-Aryan deity. The list goes on, but I trust that the point has
been made and will spare the reader further expositions on proto-Siva, pre-Aryan ‘earth
goddesses’, vedic ‘soma filters’, ‘lingams’, ‘yonis’, Vedic ‘vrajras’, and a host of other

164 The Quest for the Origins of Vedic Culture

Aryan, or non-Aryan, religious characteristics that have been reconstructed from other
Indus artifacts.

The relevance of finding items of an unambiguous Indo-Aryan nature on the Indian
subcontinent dating back prior to the second millenium b.c.e. is obvious. One report
of such an item that might have had an immediate bearing on the ethnic identity of the
Indus Valley Civilization, as well as massive repercussions for academic interpretations
of Indian protohistory and, indeed, the entire Indo-European question were it to be
accurate, somehow appeared in the pages of the Journal of Indo-European Studies (Hicks
and Anderson 1990). The authors claimed to have dated a life-sized copper-based head,
dubbed ‘Vasis^ha’s head’, with tilak markings on the brow and handlebar mustache.
The hair was styled in the manner described for the Vedic Vasisthas—coiled with a tuft
to the right. MASCA-corrected c 14 testing produced a date centered around 3700 b.c.e.
(give or take eight hundred years).

Hicks and Anderson state that the head had been tested by advanced technique in
carbon dating in Zurich by the Laboratory for Nuclear Science at the Swiss Federal
Institute of Technology, through the use of a cyclotron at the University of Califor¬
nia, as well as by means of a Davis (Pixie) ion probe and Van de Graaf linear accel¬
erator at Stanford University. The tests included spectographic analysis, X-ray dispersal
analysis, and metallography. The authors stand firm by these results (personal com¬
munication). I should note that Hicks and Anderson are not involved in the polem¬
ics of the Indigenous Aryan school; they do not contest the migration of the Indo-
Aryans into the subcontinent but feel that this find warrants backdating their arrival
to about 4000 b.c.e.

However, the origins of the head, which was “rescued from being melted down in
1958” in Delhi, are dubious. The head has not generated much attention, even among
the Indigenous Aryan school, primarily, perhaps, because it was not discovered in situ
and therefore appeared outside of an archaeological context. Therefore, although the
carbon 14 tests were determined from a small quantity of carbon deposits on the inside
surface of the cast, one could argue that this particular image had been recast from an
older copper item that had been melted down. This possibility is reinforced by the fact
that the piece bore the inscription “Narayana,” which the authors believed had been
incised at a later time. Had it been otherwise, such a discovery would have sent shock
waves through departments of ancient Indian history worldwide (although even this
probably would not have been decisive for some in terms of the Indo-Aryan identity of
the Indus Valley, since it could always have been argued that the head represented a
pre-Aryan figure whose markings and hairstyle were appropriated into Vedic and post-
Vedic culture).

Before moving on to other issues, it seems relevant to note a provocative new hy¬
pothesis suggested by Lamberg-Karlovsky (forthcoming), who draws attention to the as¬
tounding degree of cultural homogeneity in the vast area of the Indus Valley Civiliza¬
tion, juxtaposed with the lack of any evidence for a centralized political structure. Not
only is there a uniformity of culture, but the physical layout of the community is repli¬
cated irrespective of whedrer it is the 5-acre site of Allahdino or the 150 acre site of
Mohenjo-Daro. Lamberg-Karlovsky believes this “enigma” can be adequately explained
by supposing that only an exceptional social organization such as the caste system can
account for this. He finds a variety of archaeological evidence to support this. The resi-

The Indus Valley Civilization 165

dential units at Indus sites, for example, were much larger than other contemporaneous
sites, suggesting a stronger sense of kin identities or groupings. He notes that competi¬
tion in a class-structured society results in a much wider variety of styles and methods of
production, whereas in a caste system, much more uniformity is to be expected, as is
evidenced by the artifacts unearthed in the Indus Valley sites. Caste organization would
also explain the social stability of such a massive culture in the absence of a centralized
state or chieftainship. Finally, the concern with purity in a caste system is amply evi¬
denced in the archaeological record by the unparalleled attention and concern given to
the control of and access to water and sanitation; at Mohenjo Daro, there is an average
of one well for every three houses.

Lamberg-Karlovsky (forthcoming) notes that, “if valid, such a hypothesis would im¬
pose a radical rethinking of the current consensus of the allegedly Indo-European ori¬
gins of the organizational patterns characteristic of traditional Indian society” (1). For
those convinced that the Indus Valley Civilization was pre-Aryan, accepting Lamberg-
Karlovsky’s tentative hypothesis would indeed entail reconsidering the Indo-European
origins of caste. For Indigenous Aryanists, needless to say, The discovery of an Indus
Valley Civilization structured by caste would be yet another indication that this was an
Indo-Aryan culture.

The Sarasvati

The quintessential domain of the Rgveda is the land of the Sapta Sindhu, or ‘Seven
Rivers’, a land which, as we have seen, goes by this name even in Avestan sources (Hapta
Hendu). The heartland of this area more or less corresponds to the present-day Punjab
in India and Pakistan and surrounding areas. Among these seven rivers, the Sarasvati
is praised as the best 6 and as distinct in majesty (R.V. vi.61. 13). Likened to a fortress
of metal (R.V. vii. 95. 1), it presses forward like a chariot fighter going from the moun¬
tains all the way to the river (or ocean). 7 Its prestige is such that various rulers (R.V.
viii.21.18) situated themselves on its banks, and it causes die five ‘peoples’ to pros¬
per. 8 Over sixty hymns referring to Sarasvati in the Rgveda, many of which are specifi¬
cally dedicated to it, 9 attest to its importance in the world of the Vedic poets. An invo¬
cation in R.V. 10.75.5, which lists the rivers in geographically correct order from east to
west, situates Sarasvati between the Yamuna and the SutudrI (Sutlej). However, although
the other rivers in the list are all still presently extant in the north of the Indian subcon¬
tinent, nothing is to be found of the mighty Sarasvati today except for an insignificant
stream in the foothills of the Himalayas that preserves its name.

Scholars initially attempted to account for this lacuna by suggesting either that the
Indus River is the actual referent of these hymns 10 or that the verses allude to the memory
of a river outside of India, or, along with Max Muller, that the present-day Sarasvati
was once a much larger river. (See Keith and Macdonell [1912] 1967, ii, 434-436, for
a summary of opinions in this regard.) Pious Hindus, anxious to reconcile the spiritual
and physical importance accorded to the Sarasvati in the Vedic and Epic texts with its
inexplicable absence in reality, have either suggested that it was primarily a celestial
river (Sharma 1949, 53—62) or resolved the problem by concluding that it must join
the Ganga and Yamuna at the TrivenI 11 by flowing there hidden underground as the

166 The Quest for the Origins of Vedic Culture

supta nadl, or ‘Sleeping River’. (Murthy, (1980, 19.) Interestingly, their faith in the
veracity of the traditional narrative has not been placed completely in vain. Archaeologi¬
cal researches in the Cholistan desert have uncovered the bed of a once-massive river-
up to ten kilometers wide, (Misra 1989, 159)—situated between the Yamuna and the
Sudej, exactly where the Rgveda places the SarasvatT. This river is presently known as
the Hakra in Pakistan, and the Ghaggar in India. The drama of this discovery was cap¬
tured in the title of an article in Geographical Magazine: “Fabled Saraswati Flows Again.”

The first person who attempted to correlate the textual descriptions of SarasvatT with
empirical paleogeology was C. F. Oldham, in 1874. He surmised that “the waters of the
SarasvatT [are] continuous with the dry bed of a great river [Hakra], which, as local leg¬
ends assert, once flowed through the desert to the sea” (Oldham 1893, 54). Oldham
was convinced that the dry riverbed had once been fed by the Sutlej River, before the
latter changed its course westward. However, the person most responsible for drawing
public attention to the SarasvatT River was Sir Aurel Stein. 12 At eighty years of age and
almost blind, this archaeologist par excellence, adventurer, and veteran of some of the
world’s most inhospitable climes undertook an expedition in 1940-41 along the banks
of the SarasvatT/Ghaggar/Hakra River in the Bikaner area, Rajasthan, and the Bahawalpur
area, present-day Pakistan. 13

Sir Aurel was fascinated by the traditional belief that the Ghaggar or Hakra riverbed
was none other than the SarasvatT of Vedic and Epic lore and that, although long since
abandoned for most of its course, the bed once corresponded to a mighty river that
flowed down to the ocean. Since the Vedic world seemed to orient itself around the
banks of the SarasvatT, Stein felt that his expedition would be of relevance in the “still
obscure question as to where those earliest records of Indian thought were composed”
(Gupta 1989, 9). While he was successful in geographically locating this nucleus of the
Vedic world, he was forced to conclude that in terms of chronological value “it would
seem hazardous at present to use the archaeological observations concerning the an¬
cient river course for any attempt to date die references made to the SarasvatT in early
Vedic Texts” (94). It is precisely such temporal possibilities of the riverbed that are of
special relevance in the question of whether the inhabitants of the Indus Valley Civili¬
zation and the Vedic-speaking Aryans might have been one and the same.

In addition to locating and mapping a number of Harappan and post-Harappan sites
along the banks of the SarasvatT (some of which he excavated with a few exploratory
trenches), Stein, like Oldham before him, also concluded that the main reason for the
demise of the river was that the waters of its main tributary, die Sudej, had been de¬
toured, thus depriving it of the bulk of its water flow. Satellite imagery has confirmed
his observation. The present Sutlej River, rising in the Himalayas, heads in a southeast¬
erly direction directly toward the old bed of the Ghaggar. In the vicinity of modern-day
Ropar, about a hundred kilometers from where it would have coincided with the Ghaggar
bed had it continued in a straight line, it suddenly takes a sharp right-angle turn to flow
away from the Ghaggar in a westerly direction, where, after about 150 kilometers, it is
joined by the Beas. These two rivers then merge and proceed in a southwesterly course
until they join the Indus and then on to the Rann of Kutch.

The satellite imagery revealed the following data: (1) The sudden westward turn of
the Sutlej is suggestive of its diversion in the past for which no physical obstruction is
evident. 14 (2) At the point where the Sutlej would have impacted the Ghaggar riverbed

The Indus Valley Civilization 167

had it not deviated, the latter suddenly widens. Since the bed of the Ghaggar upstream
from this point is considerably narrower, this can only be explained if a major tributary
was joining the Ghaggar at this place. (3) A major paleochannel can be clearly seen to
connect the Sutlej from the point where it takes its sharp westward turn to the point on
the Ghaggar where the old bed suddenly broadens. (4) Paleochannels from the Yamuna
River show that it also once flowed into the Ghaggar and then subsequently changed its
direction three times before assuming its present course. This deviation of the Yamuna
also would have deprived the Ghaggar of a substantial supply of water (Pal et al. 1984,
492-497). There is general agreement among scholars that all this demonstrates that “it
can be stated with certainty that the present Ghaggar-Hakra is nothing but a remnant of
the RgVedic Sarasvatl which was the lifeline of the Indus Civilization” (V. N. Misra
1994, 511).

In the course of a survey project limited to only a section of the Hakra/Ghaggar in
the Cholistan desert in Bahawalpur state (representing three hundred miles of the Pa¬
kistan side of the Hakra part of the riverbed), Mughal (1993, 85) mapped out a total of
414 archaeological sites on the bed. 15 This dwarfs the number of sites so far recorded
along the entire stretch of the Indus River which number only about three dozen (Gupta
1993b, 28). The centrality of the river, both archaeologically and culturally, has led a
minority of Indian archaeologists to propose, and to begin to adopt, tire term Indus-
Sarasvati Civilization in lieu of the labels Harappan or Indus Valley Civilization.

The crucial issue in all this is the date when the Hakra/Ghaggar would have been a
full-flowing river corresponding to its state in the Rgvedic hymns. This date is seen as
powerful archaeological evidence that must be taken into consideration when dating
the composition of the Rgveda. It also has a direct bearing on the relationship between
the Indo-Aryan composers of the hymns and the Indus Valley Civilization. Mughal (1993)
proposes the following outline:

On the Pakistan side, archaeological evidence now overwhelmingly affirms that the Hakra
was a perennial river through all its course in Bahawalpur during the fourth millennium
. . . and early third millennium B.c. About the middle of the third millennium B.C.,
the water supply in the Northeastern portion of the Hakra [the Yamuna) was consider¬
ably diminished or cut off. But, abundant water in the lower (southwestern) part of this
stream was still available, apparently through a channel from the Sutlej. . . . About the
end of the second, or not later than the beginning of the first millennium B.c., the entire
course of the Hakra seems to have dried up. (4)

Mughal’s broad chronological periods are not specific enough to assist us in definitively
situating the Vedic-speaking Aryans as inhabitants of the Indus Valley Civilization. It is
significant, however, that about 80 percent of Mughal’s 414 archaeological sites along
a three-hundred-mile section of the Hakra were datable to the fourth or third millen¬
nium B.C.E, suggesting that the river was in its prime during this period. 17 The dating
range proposed by Pal et al. (1984) is no more specific: “The Ghaggar continued to be
a living river during the pre-Harappan (c. 2500-2200 b.c.e.) and the Harappan times
(c. 2200-1700 b.c.e.)” (496). A third, even wider, dating range (8,000 B.c.E.-l 800 b.c.e.)
was proposed for the SarasvatT’s channels through which it discharged into the Rann of
Kutch via the Luni River (see Ghose et al., 1979 for additional information on the
SarasvatT’s previous drainage system and course shifting). Lai (1997, 9) considers the
Sarasvatl to have been alive in Kalibangan in the third millennium b.c.e. and dried up

168 The Quest for the Origins of Vedic Culture

at the turn of the millennium: “The SarasvatT dried up around 2000 bc. This clearly
establishes that the Rigveda, which speaks of the SarasvatT as a mighty flowing river,
has to be assigned to a period prior to 2000 bc. By how many centuries it cannot be
said for certain” (Lai, forthcoming). The SarasvatT as known to the Rgveda must have
well predated the end of the second millennium b.c.e., when the entire course of the
Hakra had already dried up. A further terminus ante quern can be postulated by the
fact that Painted Gray Ware (PGW) sites dated to around 1000 b.c.e. were found on
the bed of the river, as opposed to on its banks, indicating that the river had already
become dry well before this time (Gaur 1983, 133).

Anything more than this—that is, whether the SarasvatT known to the composers of
the hymns was the river in its full glory of the fourth millennium, the more diminished
version of the third millennium, or a dwindling body of water sometime in the first half
of the second millennium b.c.e.— cannot be stated categorically. Advocates of the Aryan
invasion/migration theory can still claim that the Indo-Aryans could have arrived dur¬
ing or toward the end of the Indus civilization and then settled down on the banks of
the river. However, the Vedic hymns preserve no reference to of the river drying up
(although this is explicitly described in the Mahabharata; see 3.130.3; 6.7.47; 6.37.1 —
4; 9.34.81; 9.36.1-2). Even if the Aryans had come from outside the subcontinent, one
would have to allow at least several centuries for them to settle down on the riverbank
and completely forget their overland odyssey.

What seems apparent is that the composers of the Rgveda were living on the SarasvatT’s
banks when it was a mighty river, with no clear recollection of their having come from
anywhere else recorded in the texts they left to posterity. However, even this claim is
not without problems. Witzel (1999) notes a verse from one of the older hymns in the
Rgveda, 3.33, where mention is made of the confluence of the Sutlej and the Beas.
According to the evidence outlined previously, the Sudej was the main tributary of the
SarasvatT. Its deviation from the SarasvatT and subsequent joining with the Beas was
the principal factor that caused the SarasvatT to be deprived of much of its water input
and therefore go dry. This verse might suggest a memory of a hydronymic event that
corresponds not with SarasvatT as a mighty river but with it as a diminished stream
deprived of its principal source of water. Granted, this is a solitary reference and is dwarfed,
numerically, by the frequent references to SarasvatT’s grandeur, but it cannot be brushed
aside. If the references to SarasvatT as a mighty flowing river can be construed to suggest
that the Indo-Aryans must have been coeval with the mature Harappan phase, then the
same logic applied to this reference would suggest that the Indo-Aryans were coeval with
a Late or Post-Harappan phase (which most “Migrationists" would be prepared to con¬
sider). According to Witzel this provides “a date ad quern for this part of the RV, once
the relevant geological and geographical data have been confirmed, . . . and it speaks
against the current revisionist fashion of assigning a pre-Harappan date to the RV” (37).

Francfort (1992) of the French expedition which had been specifically studying irri¬
gation and die peopling of central Asia, has scant regard for the “mythico-religious tra¬
dition of Vedic origin, reinforced by the illusory existence of proto-historic settlements
concentrated along the banks of an immense perennial river, the ancient SarasvatT” (90—
91). In striking contrast to other dating attempts, as far as Francfort is concerned, “when
the proto-historic peoples settled in this area no large perennial river had flowed there
for a long time” (91). The Hakra/Ghaggar River, according to these researches, pre-

The Indus Valley Civilization 169

dated the entire pre- and proto-Harappan period. The Harappan sites considered by the
scholars noted earlier to have been sustained by the Hakra/Ghaggar/Sarasvatl, were,
according to the team, not actually situated on the banks of the riverbed, but were out¬
side of them, irrigated by small river channels. The team included a strong geoarchaeo-
logical element that concluded that the actual large paleocourses of the river have been
dry since the early Holocene period or even earlier (Francfort 1985, 260). Ironically, the
findings of the French team have served to reinforce the “mythico-religious tradition of
Vedic origins.” Rajaram’s reaction (1995) to the team’s much earlier date assigned to
the perennial river is that “this can only mean that the great Sarasvatl that flowed ‘from
the mountain to the sea’ must belong to a much earlier epoch, to a date well before
3000 bce”(19).

To sum up, Sarasvatl’s rediscovery, although arguably suggestive of considerable Vedic
antiquity (which one would be hard-pressed to accommodate widrin the commonly
accepted 1200 b.c.e. date for Vedic compilation), cannot be used to prove absolute
synonymity of the Indus Valley residents and the Vedic Aryans. Nonetheless, the river’s
remanifestation, albeit in the form of a paleoincarnation, has been significant in other
ways. Archaeologist S. P. Gupta (1989) voices the value that such archaeological under¬
takings can have: “At last, we found true what was recorded in oral traditions” (x). For
the geologist S. R. N. Murthy (1980), such findings are essential because “some authors
believe that it [Sarasvatl] is a ‘myth’—an imaginary river. There can hardly be a damage
equal to such interpretations due to sheer ignorance of scientific data specially which
discredit the authenticity of the Veda” (191). For many, a primary significance of the
exploration of an old, dry riverbed is that such undertakings have enhanced the episte¬
mological value of the Veda. At least in this case, the validity of sabda pramdna ‘verbal
testimony’ (recorded in written texts) has been somewhat verified by pratyaksa pramdna
‘empirical proof.

The Horse

If the Aryans cannot be identified in the archaeological record, can they at least be ex¬
cluded from it? More specifically, is the evidence from the principal archaeological cul¬
ture in the subcontinent, the Indus Valley, incompatible with the literary evidence on
the Vedic culture? As a result of the massive amount of attention the Aryan problem
has been generating in Indian academic, political, religious, and popular discourse, a
leading Indian publishing house issued a booklet written by one of the principal oppo¬
nents of the Indigenous Aryan school. The author, R. S. Sharma (1 995), concludes his
arguments with the two most common objections against those attempting to correlate
the Indus Valley with the Vedic culture: “It is claimed that the Aryans created the
Harappan culture. However, such a claim is baseless. ... It is significant that the Rg
Vedic culture was pastoral and horse-centered, while the Harappan culture was neither
horse-centered nor pastoral” (65). Likewise, Parpola (1994) states: “The view that an
early form of Indo-Aryan was spoken by the Indus people continues to have its support¬
ers. It is therefore necessary to emphasize in conclusion one important reason why the
Harappan people are unlikely to have been Indo-European- or Aryan speakers. This is
the complete absence of the horse (Equus caballus)’’ (155). Since the time of Sir John

1 70 The Quest for the Origins of Vedic Culture

Marshall, the absence of this creature has been the mainstay of the belief that the speak¬
ers of the Vedic language must have succeeded the Harappan civilization: “In the lives
of the Vedic-Aryans the horse plays an important part. ... To the people of Mohenjo-
Daro and Harappa the horse seems to have been unknown” (Marshall 1931, 111).

As we have seen the term asm ‘horse’ is a word with Indo-European credentials. In
its various forms, it is mentioned 215 times in the Rgveda and is the subject of two
complete hymns (Sharma 1995, 14). Macdonell and Keith ([1912] 1967) conclude from
one Vedic verse (RV viii. 55,3). which mentions a gift of four hundred mares, that the
animal could not have been rare in the Vedic world (I, 42). 18 There are over fifty per¬
sonalities with horse-connected names and thirty with chariot-connected names in the
Vedic literature. The horse is clearly an animal highly valued in the Vedic world. It is
perfectly reasonable to expect that if the Aryans were native to the Indus Valley their
presence would be evidenced by remains of the horse there. Such evidence, or lack thereof,
has become crucial to—and almost symbolic of—the whole Aryan controversy. The horse,
as a result, is presently “the most sought after animal in Indian archaeology” (Sharma

1974, 75).

The report claiming the earliest date for the domesticated horse in India, ca. 4500
b.c.e., comes from a find from Bagor, Rajasthan, at the base of the Aravalli Hills (Ghosh
1989a, 4). 19 In Rana Ghundai, Baluchistan, excavated by E. J. Ross, equine teeth were
reported from a pre-Harappan level (Guha and Chatterjee 1946, 315-316). 20 Interest¬
ingly, equine bones have been reported from Mahagara, near Allahabad, where six sample
absolute carbon 14 tests have given dates ranging from 2265 b.c.e. to 1480 b.c.e. (Sharma
et al. 1980, 220-221). 21 Even more significantly, horse bones from the Neolithic site
Hallur in Karnataka (1 500-1300 b.c.e.) have also been identified by the archaeozoolo-
gist K. R. Alut (1971, 123). 22 These findings of the domestic horse from Mahagara in
the east, and Hallur in the south, are significant because they would seem inconsistent
with the axiom that the Aryans introduced the domesticated horse into the Northwest
of the subcontinent in the later part of the second millennium b.c.e. Due to the contro¬
versy generated by Alur’s report, a reexcavation of Hallur was undertaken to collect fresh
samples of animal bones. Alut (1992) again insisted that specimens of Equus caballus
Linn were definitely present in the collection. His response is worth quoting at length
to give a sense of the controversy and significance surrounding this animal:

When I wrote this report, I least expected that it might spark off a controversy and land
me in the witness box before the Indian historians’ jury. ... I was apprised of the gravity
of the situation when I began to get letters asking me for clarification of the situation
against the prevalent belief that the horse is a non-indigenous species and was introduced
into India only by the (invading) Aryans. . . . To make my position clear, I wrote in my
article . . . that whatever may be the opinion expressed by archaeologists, it cannot either
deny or alter the find of a scientific fact that the horse was present at Hallur before the
(presumed) period of Aryan invasion. ... 1 have only declared the findings that horse
bones were traced in the faunal collection from Hallur and am responsible to that extent
only. (562, italics in original)

In the Indus Valley and its environs, Sewell and Guha, as early as 1931, had re¬
ported the existence of the true horse, Equus caballus Linn from Mohenjo-Daro itself, 23
and Bholanath (1963) reported the same from Harappa, Ropar, and Lothal. Even
Mortimer Wheeler (1953) identified a horse figurine and accepted that “it is likely enough

The Indus Valley Civilization 171

that camel, horse and ass were in fact all a familiar feature of the Indus caravan” (92).
Another early evidence of the horse in the Indus Valley was reported by Mackay, in
1938, who identified a clay model of the animal at Mohenjo-Daro. Piggott (1952, 126,
130) reports a horse figurine from Periano Ghundai in the Indus Valley, dated some¬
where between Early Dynastic and Akkadian times. Bones from Harappa, previously
thought to have belonged to the domestic ass, have been reportedly critically reexam¬
ined and attributed to a small horse (Sharma 1992-93, 31). 24 Additional evidence of
the horse in the form of bones, teeth, or figurines has been reported in other Indus
sites such as Kalibangan (Sharma 1992-93, 31 ); 25 Lothal (Rao 1979), 26 Surkotada
(Sharma 1974), and Malvan (Sharma 1992-93, 32). 27 Other later sites include die Swat
Valley (Stacul 1969); Gumla (Sankalia 1974, 330); Pirak (Jarrige 1985); 28 Kuntasi (Sharma
1995, 24); 29 and Rangpur (Rao 1979, 219). A. K. Sharma (1992-93) comments on the
academic reaction to these not inconsiderable reports:

It is really strange that no notice was taken by archaeologists of these vital findings, and
the off-repeated theory that the true domesticated horse was not known to the Harappans
continued to be harped upon, coolly ignoring these findings to help our so-called veteran
historians and archaeologists of Wheeler’s generation to formulate and propagate their
theory of “Aryan invasion of India on horse-back.” (31)

The exact species of the equid is the crucial issue in these identifications. The debate
over horse bones has become acrimonious ever since Zeuner (1963) questioned the iden¬
tification of Ross’s pre-Harappan findings: “The earliest horse remains so far reported
come from Rana Ghundai in Northern Baluchistan . . . the date of which is regarded as
earlier than 3000 B.c. . . . this identification cannot be accepted as reliable unless it is
carefully checked” (332). Since this challenge, detractors of the Indigenous Aryan school
have been able to reject claims of horse bone findings as unreliable, since the bones
might have appertained to the domestic ass, Equus asinus, or the hemione, Equus hemionus
khur. Although the latter is indigenous to the Northwest of the subcontinent it is Equus
caballus that is the sought-after Aryan steed. Until recently, these distinctions had ham¬
pered widespread acceptance of any existence of the horse at all in the Indus Valley
because there are only minor differentiating features between the various species of Equus
(See Meadow 1987, 909). These are either difficult for experts to identify or, unless
the specific distinguishing parts of the skeleton are found (certain teeth and the phalan¬
ges—toe bones—are particularly important for differentiating equid subtypes), impossible
to determine with certainty. Many of the remains could have belonged to either Equus
caballus Linn or to some other member of the horse family and are thus rejected as
incontestable evidence of the former. Thus Meadow (1987) writes:

There are, as yet, no convincing reports of horse remains from archaeological sites in
South Asia before the end of the second millennium bc. Many claims have been made
(e.g., Sewell 19.31; Nath 1962, 1968; Sharma 1974) but few have been documented with
sufficient measurements, drawings, and photographs to permit other analysts to judge for
themselves. An additional complication is that some specimens come from archaeologi¬
cal deposits which could be considerably younger than the main body of material at the
site. (908)

The situation took a new turn, somewhat melodramatically, a few years ago. The
material involved had been excavated in Surkotada in 1974 by J. P Joshi, and A. K.

172 The Quest for the Origins of Vedic Culture

Sharma subsequently reported the identification of horse bones from all levels of this
site (circa 2100-1700 B.C.E.). In addition to bones from E quits asinus and E quus hemionus
khur, Sharma reported the existence of incisor and molar teeth, various phalanges, and
other bones from E quus caballus Linn (Sharma 1974, 76). Although some scholars
accepted the report, doubts about the exact species of E quus represented by the bones
prevented widespread recognition of Sharma’s claim. Meadow (1987) has written: “It is
on the basis of this phalanx that one can ascertain from die published photographs that
the ‘horse’ of Surkotada, a Harappan period site in the little Rann of Kutch, ... is
likewise almost certainly a half-ass, albeit a large one” (909).

Twenty years later, at die podium during the inauguration of the Indian Archaeo¬
logical Society’s annual meeting, it was announced that Sandor Bokonyi, a Hungarian
archaeologist and one of the world’s leading horse specialists, who happened to be passing
through Delhi after a conference, had verified that the bones were, indeed, of the do¬
mesticated Equus caballus: “The occurrence of true horse (Equus caballus L.) was evi¬
denced by the enamel pattern of the upper and lower cheek and teeth and by the size
and form of incisors and phalanges. Since no wild horses lived in India in post-pleistocene
times, the domestic nature of the Surkotada horses is undoubtful (reproduced in Gupta
1993b, 162; and Lai 1997, 285). Sharma, vindicated, received two minutes of applause
from the entire assembly (Sharma 1992-93, 30); there now seemed to be no doubt
about the horse at Surkotada. Sharma comments on this validation:

This was the saddest day for me as the thought flashed in my mind that my findings had
to wait two decades for recognition, until a man from another continent came, examined
the material and declared that “Sharma was right.” When will we imbibe intellectual
courage not to look across borders for approval? The historians are still worse, they feel
it is an attempt on the part of the “rightists” to prove that the Aryans did not come to
India from outside her boundaries. (30)

This poignant statement reveals two significant dimensions to the Aryan problem.
Sharma's comments afford us a glimpse at the political tension underlying even as in¬
nocuous a piece of data as a horse’s molar. Second, and (only) partly as a corollary of
the emotionalism that the Aryan problem generates in India, many Indian scholars still
value the opinion of a Western scholar more than that of their compatriots. 30

Bokonyi’s endorsement of the Surkotada findings have also been challenged, in turn
however. Bokonyi had identified six tooth specimens that could “in all probability be
considered remnants of true horses” (1997, 298-299). Meadow’s subsequent investiga¬
tions into these identifications caused him to conclude that “we agreed to disagree on
all these matters and noted the need for further research. ... we cannot accept without
serious reservations Bokonyi’s identifications of any of the Surkotada material as true
horse, but in the end that may be a matter of emphasis and opinions” (1997, 315). 31
Unfortunately, Bokonyi was not able to write a reply before his death. Meadow’s reser¬
vations refer to the problems outlined previously relating to the difficulty of distinguish¬
ing between the different equid species in an unambiguous fashion.

Although A. K. Sharma claims that the bones of E quus caballus have been discov¬
ered “from so many Harappan sites and that too right from the lowest levels [thus es¬
tablishing] that the true domesticated horse was very much in use by the Harappans”
(1992-93, 33), with the exception of the report from Rana Ghundai, which was ques-

The Indus Valley Civilization I 73

tioned by Zimmer, 32 and Piggott’s reported horse figurine from Periano Ghundai, it
would appear that much, if not all, of even the contested evidence comes from strata
associated with later Harappan sites or at least not from the Pre-Harappan or Early
Harappan period. This, of course, as with the SarasvatT, and the fire altars at Kalibangan,
leaves scope for the proposal that, even if for argument’s sake one is prepared to allow
the claims regarding horse bones, it could still be argued that the Aryans could have
introduced the “true” horse into the subcontinent during the Harappan period itself:
“Indeed with the present state of evidence it would be unwise to conclude that there is
any proof of the regular use of the horse in pre-Harappan or Harappan times” (Allchin
and Allchin 1982, 191). Even B. B. Lai (1997), who is prepared to question the theory
of Indo-Aryan migrations, has to acknowledge that “one would like to have much more
evidence, to be able to say that the horse did play a significant role in the Harappan
economy” (162).

A more significant horse lacuna, in the opinion of some, is that “several animals
appear on Harappan seals . . . but the horse is absent” (Sharma 1995, 17). 33 In view of
the fact that thousands of seals have been found, this absence is quite remarkable and
potentially fatal to the Indigenous Aryanists, since, if the Aryan horse were indeed present
in the Indus Valley, surely it would have attracted the artistic attention of, at least, the
odd seal maker or two. Parpola, critiquing Sethna’s rejection of the Aryan Invasion
hypothesis, drew attention, among other things, to the horse lacuna on the seals. How¬
ever, the Indigenous Aryanists are extremely resilient to what might appear to be fatal
criticisms from their opponents. Sethna (1992) counters:

As there are no depictions of the cow, in contrast to the pictures of the bull, which are
abundant, should we conclude that Harappa and Mohenjo-daro had only bulls? And what
about that mythical animal, the unicorn, which is the most common pictorial motif on
the seals? Was the unicorn a common animal of the proto-historic Indus Valley? Surely,
the presence or absence of depictions cannot point unequivocally to the animals known
and decide for or against Aryanism. (180)

Unless we are to suppose the existence of unique, monogenetic species of Harappan
bull, it must be conceded that Sethna has a point; the cow is, indeed, also completely
absent from seal depiction, although massive quantities of bovine bones have been found.
Actually, such arguments were dismissed as early as Sir John Marshall’s time: “The
negative argument... is not altogether conclusive; for the camel, too, is unrepresented,
though the discovery of a bone of this beast . . . leaves little doubt that it was known”
(Marshall 1931,28). 34 It seems safer to assume that certain animals appear on the seals
at the exclusion of others as a result of culturally conditioned criteria, rather than be¬
cause they document the complete zoological diversity of the Harappan landscape.

It is fair to note that even if some or all of the identifications of horse evidence is
ambiguous, this should not then be translated into proclamations that the horse was
not present in the Indus Valley; one would only have a right to say that the findings
could belong either to Equus caballus or to another type of equid and that there has so
far been no unambiguous evidence. The result of Bokonyi’s endorsement of the horse
bones at Surkotada was that “many other excavators of Harappan sites started search¬
ing for the bones of Equus caballus, with at least one further siting claimed at Dholavira"
(Sharma 1992-93, 31). Obviously, the horse would never have been an issue had it not

174 The Quest for the Origins of Vedic Culture

been linked with the Indo-Aryans. If Indigenous Aryanists seem keen to promote any
reports of horses in the Harappan civilization as evidence of the Aryan presence, their
detractors seem just as keen to find reason to challenge all such reports. Clearly, were
it not for the politicization of the Aryan issue, the reports of horse evidence, albeit sparse,
would hardly have raised any eyebrows.

There is a strong feeling among many of the archaeologists I interviewed that there
are probably horse bones present in the many bags of bones that can be seen lying
around in any Indian museum, but that they had never formerly been checked properly
because the Indigenous Aryan position has only relatively recently climbed to ascen¬
dancy in India. In other words, scholars had previously just assumed that the Aryans
were invaders and therefore were not so concerned about identifying horse bones. Hence,
A. K. Sharma (1992-93) issues an appeal to young, up-and-coming archaeologists to be
very attentive in their handling of animal bones and skeleton remains found in excava¬
tions, and not to attempt any personal weeding out of the material in order “to lessen
the volume of finds, for fear of cost and labour involved in transporting them to head¬
quarters” . . . [Once here,] they should not be . . . dumped ... in some packing case for
decades” (34). 35

This is an important petition because unless horse bones are undeniably found in
the Early, Pre-, and Mature Harappan strata, the Indo-Aryan speakers may be (and al¬
ready are) allowed a degree of synthesis with the later Harappan civilization, but their
status as intruders, albeit considerably earlier than previously held, will still not be con¬
sidered convincingly undermined to die satisfaction of all. As two scholars who reviewed
the horse evidence conclude, “considering that the presence of the horse during the
Harappan period is a matter of popular controversy in Indian archaeology, the subject
deserves more serious and systematic treatment than it has so far received” (Thomas
and Joglekar 1994, 187).

It might be timely to again briefly refer to some of the Indigenous Aryan positions
on this crucial issue, which were discussed in chapter 6. The horse has always been an
elite animal in the subcontinent, but a nonnative and rare one, hence the paucity of
evidence in the archaeological record. This, however, need have nothing to do with the
indigenousness of the Indo-Aryans themselves, who simply imported this prized ani¬
mal in the ancient period, as they have always done right up to the modern period.
Bridget Allchin (1977) voices a similar position after the exchanges between Bokonyi
and Meadow:

From early historical times forward we know that horses have been regularly imported to
South Asia. We also know the Indus had a long tradition of trade with centres to the west
and north. Would it be surprising therefore if horses were occasionally acquired through
trade, ultimately reaching the Indus world by land or sea? This would account for the occur¬
rence of a small number of their bones in various contexts without the need to assume their
presence must necessarily be associated with profound cultural change. (316)

The paucity of horse bones could simply denote the possibility that die animal was an
elite and rare item. All this, of course, will be considered special pleading by the detrac¬
tors of the Indgenous school.

Another observation that needs to be pointed out is that a number of scholars are
prepared to consider that the Bactria Margiana Archaeological Complex (BMAC), which

The Indus Valley Civilization l 75

will be discussed in the next chapter, is an Indo-Aryan culture. The horse has been
evidenced in this culture in the form of representations in grave goods. However, no
horse bones have been found despite the availability of a large number of animal bones.
This again underscores the point that lack of horse bones does not equal the absence of
horse. Nor, at least in the opinion of those who subscribe to the Indo-Aryan identifica¬
tion of the BMAC, does this lack equal the absence of Indo-Aryans. Therefore, anyone
prepared to associate the BMAC culture with the Indo-Aryans cannot then turn around
and reject such an identification for the Indus Valley on the grounds of lack of horse
bones in the latter.

As a final note, if the horse is to be promoted as the Achilles’ heel of the Indigenous
Aryan school, those advocating the Dravidian speakers as the inhabitants of the Indus
Valley have their own lacuna to account for:

It has been stated by the supporters of the Dravidian theory that the Aryan invaders chased
away the Dravidian-speaking Harappans to the southern part of India. . . . Those who
hold this view have squarely to answer: If the Aryans pushed the Harappans all the way
down to South India, how come there are no Harappan sites at all in that region? The
southernmost limit of the Harappan regime is the upper reaches of the Godavari. There
is no Harappan site south of that. (Lai 1997, 284)

This observation merits consideration. Even if the Dravidians had not been pushed
down, but subsumed, one would have some grounds to expect seals, samples of the
Indus script, or any item of material culmre from any hypothetical Dravidian Harappans
to have been shared with their fellow Dravidian speakers down South (and consequently
surface in the archaeological record). This is not the case. Even before the Harappan
decline, one would have expected much greater trade and cultural exchange with the
South had the two areas shared a common language and culture. The same can be said
for any attempts to link the Harappans with the Munda speakers from the east. The
Dravidian Harappanists, while very conscious of the horse evidence, tend to overlook
this objection to their own position. On the other hand, there are plenty of examples of
different types of socioeconomic cultures and civilizations that nonetheless share the
same language: Lai himself, in a later chapter, argues that the nomadic Indo-Aryan
composers of the Rgveda, could have coexisted with urban Indo-Aryans of the Indus
Civilization.

The Chariot

The spoked-wheel chariot also is fundamental to Aryan identification. If the Aryans
were a principal linguistic community within the Indus Valley Civilization, their exist¬
ence there should be confirmed not only by the horse but also by the spoked-wheel
chariot. This piece of technology, called ratha in Sanskrit (< PIE *rota), is common to
the Indo-European peoples, since, like the horse, its nomenclature has cognates in Indo-
Iranian, Italic, Celtic, Baltic, and Germanic. Likewise, the terms for the parts of the
chariot—the wagon pole, harness, yoke, and wheel nave—also have cognates generously
distributed in various Indo-European languages. Either the Proto-IndcvEuropeans knew
the chariot or it was a later innovation that swept across the Indo-European-speaking

176 The Quest for the Origins of Vedic Culture

area. In either event, the Indo-Aryans certainly utilized the technology after the disper¬
sion of the various Indo-European tribes, an inference evidenced by the central role it
plays in the Veda (which parallels its importance in other old Indo-European texts such
as the Homeric hymns). The quest for the Indo-Aryans, then, as a result of a logic analo¬
gous to that impelling archaeologists to look for the horse, unavoidably involves search¬
ing for archaeological evidence of the chariot.

Although iconographic representations of solid-wheeled vehicles are attested as early
as the fourth millennium b.c.e ., 37 the earliest evidence of the spoked-wheel chariot oc¬
curs in wheel imprints in the Sintashta cemetery dating to about 2000-1800 b.c.e., in
representations from Syrian seals from the Anatolia, Uruk, and on eighteenth and sev¬
enteenth centuries b.c.e. Syrian seals (Mallory 1989, 69). Just as Equus caballus Linn,
is the precise equid scholars have selected to demonstrate Aryan identity, the spoked-
wheel chariot is the specific vehicle involved in the same task. Scholars have generally
held that the horse-drawn, spoke-wheeled chariot was introduced into the Near East by
Indo-Aryan-speaking peoples intruding from the north. The principal support for this
hypothesis is the famous Hittite manual from Bogazkoy, mentioned previously, wherein
technical terms relating to the training of chariot horses are used. The text is written in
Hittite, an Indo-European language, but, as 1 have discussed, the technical terms are in
a dialect of Indo-Aryan very closely connected with Vedic Sanskrit. Although Piggott
(1977, 1983) maintained that the arrival of these Indo-European groups did, indeed,
inaugurate the new technology in Mesopotamia, 38 Littauer and Crouwel (1979, 68)
prefered to consider the chariot a local, Mesopotamian development, 39 while Moorey
(1986, 211) considered that a combination of local and external factors resulted in the
innovation.

Evidence for chariot use prior to the common era is documented much less by actual
archaeological finds of the vehicle itself than by iconographic and literary evidence of it.
In India, however, the earliest evidence of the chariot seems to have been at Atranjlkhera
in the Upper Gartga Basin sometime between 350 and 50 b.c.e. (Gaur 1983, 373) and
representations occurring in the late first millennium, on stupas, Ashokan pillars and
Kushana art. 40 There is plenty of evidence of wheeled vehicles in the Indus Valley,
particularly in the form of miniature models or toys (Mackay 1943, 162-166), 41 but
nothing suggestive of spoked-wheels or chariots. Of course, this is negative evidence
based on argumentum ex silentio, and one could argue that it is not practical to construct
miniature spoked wheels from clay hardy enough to withstand the abuse of children,
hence the lacuna of the spoked wheel. Moreover, many of the toys are missing their
wheels, so spoked wheels could have coexisted with solid ones, as is the real-life case in
the Sindh today. Piggott (1970, 202), at least, was not averse to considering that some
of the wheels that are missing from the carts may have been spoked and correctable
with intruding Aryans. Nonetheless, the absence of definite evidence of the spoked wheel
is a lacuna that is levied against the case of the Indigenous Aryanists.

This absence is mitigated to a certain degree in that, even if we accept the latest date
assigned by scholars to the Rgveda, namely, 1200 b.c.e., the chariot as known to this
text must have unquestionably been in existence on the subcontinent for approximately
a millennium before becoming evidenced in iconographic form just before the com¬
mon era. Witzel’s date of 1 700 b.c.e. increases that period to a millennium and a half.
If we accept the opinion of those who hold that the “fire altars” in Kalibangan, the

The Indus Valley Civilization 111

evidence of the horse in later Harappan sites, and the circumstantial evidence of the
Sarasvatl suggest (even to scholars who insist on their ultimate external origins) that the
Indo-Aryans had some presence at least in the later period of the Harappan civilization,
the period increases further still. The archaeological argumentum ex silentio clearly shows
its limitations in this period during which we know the chariot was extant from the
literary evidence, despite the fact that it has not been verified archaeologically. Obvi¬
ously, the farther back in time we go, the more the likelihood of finding such icono-
graphic evidence decreases. Indeed early archaeological evidence for the oldest discov¬
ered spoked chariot wheel in the Sintashta Cemetery of the Andranovo culture, can
only be inferred from soil discoloration from die decayed spokes, and slots in grave
floors (Piggott 1983, 91).

Therefore, since archaeology has not unearthed any signs of spoked wheels or chari¬
ots in the period of one, or one and a half, millennium during which we know from
Vedic texts that the chariot was very much present in the Northwest of the subconti¬
nent, it seems legitimate to question the validity of this evidence as the final arbitrator
of Aryan origins, especially since chariots leave little archaeolgical residue. In the event
of a lucky find, archaeology can confirm, in such cases, but it cannot deny, in the ab¬
sence of the same. Such limitations causes some scholars to lament the perceived
overdependence on archaeological evidence: “Unfortunately some archaeologists seem
to have no eyes to see anything but what the spade digs up from the bowels of the earth,
and no ears to hear anything that is not echoed from excavated ruins” (Majumdar 1959,
11). Nonetheless, if the method of archaeology is to be allowed any value at all, then the
burden of proof rests with those who wish to argue that chariots (and horses) were in
use in the Indus Valley Civilization. While any archaeologist would agree that “absence
of evidence is not evidence of absence,” theories can only be established on the pres¬
ence or absence of evidence until falsified by later discoveries. Otherwise, archaeologists
would have to accept everything as possible and be unable to formulate any hypotheses
at all concerning cultural evolution.

The Indus Script

While some of the evidence discussed here has caused many South Asian archaeolo¬
gists to reconsider or modify their positions regarding Indo-Aryan intrusions into the
subcontinent, one piece of evidence that has been in academic custody since even be¬
fore the official discovery of Harappa could, if made to testify, immediately bear witness
to the linguistic identity of the inhabitants of the Harappan civilization. This would
have immediate repercussions for the entire Indo-Aryan debate and, indeed, to all in¬
tents and purpose could conclude the whole matter. The Indus seals (64), and a few
other assorted items are imprinted with symbols of an unknown script. The decipher¬
ment of this script would determine in one decisive stroke whether the principal lan¬
guage spoken in the Indus Valley was Indo-Aryan, Dravidian, Munda, or some other
linguistic entity. Despite dedicated efforts spanning almost seventy years (which have
included the use of sophisticated computer techniques), the script remains tantalizingly
resistant to being deciphered. Needless to say, if these endeavors ever bear fruit, the
need for searching through old bags of bones in quest of Equus cahallus molars will be

1 78 The Quest for the Origins of Vedic Culture

eliminated in one decisive stroke. If the script can be shown to be some form of Indo-
Aryan, there can be. no further serious debate about the dominant linguistic identity of
the Indus Valley. On the odier hand, if it can be shown to be a non-Indo-Aryan lan¬
guage, then the entire case of the Indigenous Aryanists will lose most of its cogency.

Just as this book is going to press, the discovery of what appears to be proto-Indus
writing on shards of pottery from the Ravi phase that are as early as 3500 b.c.e. has
been announced. 42 If this script contains the same language as the later Indus script
(there is always die possibility that the same script might contain different langauges at
different chronological periods), then any Indo-Aryan decipherment would have major
repercussions for the whole Indo-European problem, not just the identity of the Indus
Valley. The Indo-Europeans could still have been an undivided entity at this time, as
will be discussed in chapter 12. An Indo-Aryan language during this period will drasti¬
cally affect the whole Indo-European homeland locating quest and provide unprecedented
support for Indian homeland proponents. The script, for a number of reasons, is of
great interest to Indigenists and Migrationists alike.

The fact that both the language and the script are unknown makes the task extremely
difficult. Other decipherment attempts have involved known languages in an unknown
script (when decipherment is virtually guaranteed) 43 and unknown languages in a known
script—a much more difficult combination. Although Hittite is an example of an un¬
known language successfully extracted from a known script (cuneiform), this second
category is by no means guaranteed to bring success. Etruscan continues to refuse to
yield to complete translation attempts despite being written in a form of Roman script,
with even some bilingual inscriptions available in Latin. All this should sober any would-
be decipherer intent on attempting the third and most difficult type of decipherment
project: an unknown language in an unknown script. Such is the Harappan case. The
task is not impossible: cuneiform, Linear B, and the Egyptian hieroglyphics are success¬
ful examples of previously unknown languages being extracted from unknown scripts,
but, without bilingual inscriptions, success in this category is extremely unlikely. Cunei¬
form and the Egyptian hieroglyphs were deciphered with the aid of lengthy bilingual
and trilingual inscriptions containing known languages. Even then, Jean-Francois
Champillon, who succeeded in deciphering the hieroglyphs, spent fifteen years compil¬
ing all the signs and variants before even attempting a reading, despite having access to
the famous Rosetta stone, where the Egyptian pictographs were accompanied by a Greek
translation. The decipherment of the Linear B script without any bilingual inscriptions
by Michael Ventris in 1951 ranks as a great and rare achievement in the annals of
decipherment.

Like Linear B, there are no bilingual inscriptions of die Indus script, but, unlike
Linear B, the Indus seal decipherment is further hampered by the extreme brevity of
the seals, each of which contains an average of only five symbols. The longest inscrip¬
tion is twenty-six symbols on three sides of an amulet, while the longest single-sided
inscription is only seventeen symbols. To extrapolate a morphologically or syntactically
consistent language from seals averaging a mere five signs is a nearly impossible task.
The length of the seals parallels that of the Etruscan inscriptions, the brevity of which
has also prevented their decipherment (apart from a few names). Yet, in the Etruscan
case, some bilingual inscriptions are available, as is an extensive knowledge of the cul¬
tural setting of the people.

The Indus Valley Civilization I 79

The brevity of the Indus inscriptions allows various sounds from the same, or differ¬
ent, languages to be assigned to the same symbols and, with a little fudging and coax¬
ing, still result in meaningful words on some kind of a regular basis; hence the plethora
of so-called decipherments. As Allan Keislar (n.d., 6) notes, however, spurred on by
apparent success in assigning phonemes to some of the symbols and producing mean¬
ings that appear reasonable, at least to themselves, many would-be decipherers believe
the reality of their system has been validated. If even a single lengthy text were to be
found, then all the competing sound values could be put to the test to see if a seman¬
tically coherent and morphologically and syntactically consistent language emerged.
Ventris applied his tentative sound correspondences to the Linear B script and was
surprised to find Greek emerging from the texts (he was expecting a pre-Greek non-
Indo-European language). Likewise, the BrahmT inscriptions were decipherable from a
few Ashokan pillars, but the inscriptions were lengthy enough to allow a clearly identi¬
fiable Prakritic Sanskrit to emerge.

One of the few things on which most scholars do agree is the direction of the writing
on the seals, which is obviously the first feature to be established before any attempts
can be made to “read” the signs. Lai, in 1966, pointed out that on some inscriptions
incised on pottery shards, certain symbols had been superimposed over the symbol to
their right, which they partly effaced, thereby indicating that they had been inscribed
after the right-hand symbol had been written; this would indicate writing from right to
left. An inscription along three sides of another seal also shows that the writing could
only have been from right to left. However, the situation is complicated somewhat by
similar evidence indicating that the writing was sometimes incised in a boustrophedonic
fashion (‘as the ox plows,’ i.e., left to right to left, etc.).

There is still no agreement among scholars regarding whether the script is logo-syl¬
labic, or syllabic. Many scholars feel that it represents a logo-syllabic script, based on
the fact that they identify approximately four hundred distinct principal signs (exclud¬
ing hundreds of additional signs, which they construe either as variations of these or as
conjunct signs). Logographic scripts, where a sign denotes a complete word, need a much
larger number of signs—up to nine hundred—to represent the range of words required
in communication. Syllabic scripts, where each sign represents a syllable, need fewer
signs—typically fifty to one hundred—to denote the entire possible range of syllables utilized
in a language. 44 Since the number of distinct Indus signs appears to fall between these
two categories, the script has been designated as logosyllabic by some. In the opinion of
other scholars, however, the script is considered to be primarily syllabic. Such scholars
perceive a much smaller number of basic signs in the script and construe the remainder
as being conjunct forms or voweled variants of these basic signs.

Sanskrit Decipherments

Several scholars have a priori ruled out the possibility of a Sanskrit language being
uncovered in the seals. Zide (1979) feels that this is “completely inadmissible on the
grounds of chronological incongruity . . . and so is immediately discredited” (257). In
other words, since the Indo-Aryan language is assumed not to have entered the subcon¬
tinent until after the Indus Valley, it is not even a candidate. Along the same lines,
Possehl (1996) states that “Indo-Aryan is dismissed since the Fairservis position is that

180 The Quest for the Origins of Vedic Culture

representative languages of the family arrived in the subcontinent at ca. 1500 B.C.”
(153). Such a priori grounds will not be invoked in this analysis, which has suspended
all presuppositions concerning the origins of the Indo-Aryans.

However, other scholars believe that the script cannot be Sanskrit for structural rea¬
sons. Computer analyses of the morphology of the script suggest to them that it did not
have prefixes and inflectional endings. If this were to be proved true, then Indo-Aryan
and any other early Indo-European language, which do contain such features, would
have to be disqualified as likely candidates. This would result in the entire Indigenous
Aryan position losing cogency. It is imperative for the entire Indigenous case that the
script be Indo-Aryan. On the other hand, it has also been argued that it is unlikely that
the script is Dravidian, since it uses a numbering system with a base ten. Dravidian
uses base eight.

S. R. Rao’s claim of decipherment, which he has presented in a variety of publica¬
tions (1979, 1991), is the best known attempt from the Indigenous Aryan school be¬
cause of this scholar’s preeminent status in Indian archaeology. An official Indian gov¬
ernment tourist publication on Dvaraka presents one of Rao’s translations for a seal
found off Dvaraka as factual (Keislar n.d., 19), 45 Since his efforts are often promoted by
the Indigenous Aryan camp, Rao’s work merits some attention.

Based on the strata in which certain seals were found, Rao (1991) proposes that the
script underwent an evolution from logosyllabic, in the Early Harappan period, to pho¬
netic, in the Mature and Late Harappan period. Rao identifies thirty-four basic cursive
signs that occurred in the Mature Harappan period, but, since he considers some of
these to be alternative signs, he reduces the number occurring in the Late Harappan
period to twenty-four. Thus the script, in Rao’s system, is essentially phonetic, although
it developed from a logosyllabic progenitor. He sees all the remaining Indus signs that
are not pictographic or basic as being either compound forms, diacritic markers, or vocalic
indicators that qualify this basic group.

Having reduced the total number of signs to this essential core, Rao’s method in¬
volves comparing this nucleus with similar signs from the Semitic script, specifically the
Old North Semitic and South Semitic scripts of the Phoenicians, Hebrews, and South
Arabs. The oldest inscriptions in these scripts date from 1600 to 1200 B.C.E., a period
that overlaps with the Late Harappan period and the latest attested samples of the Indus
script. Rao considers this method justifiable because there were substantial trading and
cultural connections between the Indus Valley and Mesopotamia. He finds seventeen
of his thirty-four signs common to the two scripts, with some minor alterations, and
assigns the same phonetic value from these seventeen Semitic signs to their Indus coun¬
terparts. He also finds 16 of the basic Indus signs have a close similarity to the much
later South Asian BrahmT script, despite the chronological gap of circa one millennium
between the two. 46 His readings, however, seem to be based mostly on the letters with
Semitic similarities. Rao finds that the seals read by his method reveal an Old Indo-
Aryan language.

While Rao’s procedure, so far, is a reasonable one to consider, at least as a working
hypothesis, his treatment of two other signs from his basic core, the ‘man’ sign and the
‘fish’ sign, which are among the most common of the Indus signs but do not have
Semitic equivalents, seems much more arbitrary. Taking the Sanskrit word for ‘man’
nr/nara and for ‘fish’ sakala, Rao assigns the second letter r from the former and the

The Indus Valley Civilization 181

first letter s' from the latter to the man and fish signs, respectively. He follows a similar
procedure in assigning sound values to various other common pictographic signs, ex¬
cept, in these instances, instead of extracting a single phoneme, he chooses to assign a
full syllable from the Sanskrit word corresponding to the sign (i.e., sak for the ‘bird’
picture). Rao justifies such apparent arbitrariness, including the choice of the particular
Sanskrit synonym he decides to select for the picture (i.e., sakuna ‘a bird of omen’ as
opposed to the more common words for bird such as paksi or vihaga ) simply because it
works—words meaningful to his sensibilities are produced by the assignment of certain
syllables as opposed to others. 47 The problem, of course, is that the script can also be
made to “work,” at least to the satisfaction of other would-be translators, by assigning a
wide variety of syllables to the same symbols. Keisler produced a chart showing the
wide assortment of phonemes or syllables assigned to the same Indus sign by the vari¬
ous principal decipherment contenders—all of whom seem satisfied that their particular
versions are valid because they work.

Another interesting decipherment attempt by Subhash Kak, a professor in the elec¬
trical and computer engineering department of Louisiana State University, also connects
the Indus script with Indo-Aryan and is noteworthy because of its innovative method¬
ology. As Possehl (1996) notes, “Kak’s work brings to it the serious mind of a scientist,
and his ability to deal with the problem in a quantitative way is needed.” As he points
out, this is a welcome change: “Kak’s great strength is that he brings mathematics into
play on the study of the script. His is not the old-fashioned iconography analysis, tried
again and again” (148). Kak sets out to determine whether the later BrahmT script might
be connected with the Indus signs, since it is the oldest known orthographic script for
the Indie languages. Like Rao, Kak considers the script to be syllabic on the grounds
that if the thirty-three consonants known to BrahmT are conceived of as being modifi¬
able in ten different ways by the addition of ten different vowel signs, the resulting number
approaches the basic corpus of signs recognized by most scholars. Kak reduces the 300-
odd Indus characters down to a set of 39 primary signs, most of which he feels are
easily identifiable (either as primary signs or as combinations of primary signs). These
39 signs, singly or in combination, account for 80 percent of all the signs; in compari¬
son, BrahmT has 45 primary letters.

Kak’s next observation is that although logosyllabic scripts have more signs, each
individual sign occurs more consistently than in a syllabic system where some signs
occur with high frequency and others rarely. Thus, from a random 13,000 letters in
English, a syllabic script, for example, one would expect to see E about 1,700 times, T
about 1,200 times, and J, Q, X, and Z only 17, 16, 20, and 10 times each, respectively.
In parallel fashion, from a corpus of 1 3,000 Indus signs, two signs constitute 2,000 of
the total, while more than 200 signs occur less than 10 times each (Kak 1987a, 54).
Kak then takes ten different Sanskrit texts of one thousand words each (in order to
represent a range of different genres), processes the combined ten thousand words through
a computer, and draws up tables of the ten most commonly occurring Sanskrit pho¬
nemes. These are (in decreasing order): t, r, v, n, m, y, s, d, p, and k. He then assigns the
BrahmT characters for these letters and compares them with a table of the ten most
commonly occurring Indus symbols. He finds there are convincing parallels between
four of the two sets, BrahmT v, m, t, and s, which even appear in the same order of
frequency. Certain changes, he feels, have taken place (e.g., the fish sign had been flipped

182 The Quest for the Origins of Vedic Culture

sideways), but he argues that such modifications can also be seen in the evolution of
BrahmT to Nagarl. According to Kak’s calculations (1988, 135), the probability of these
resemblances occurring by chance is 0.1 x 10 -12 . It should be noted, however, that the
phonemes he assigns to the Indus script on the basis of their similarities differ from the
phonemes Rao assigns to his BrahmT/Indus correlations.

Nonetheless, granting all of Kak’s observations, all that can be inferred at this point
is that perhaps the BrahmT script was derived from the Indus script. This possibility
does not deal with the language family represented by the script. Kak is wise enough
not to proclaim that he has deciphered the script or to produce long lists of transla¬
tions, but he does offer one tentative, but potentially very significant, morphological
observation. He sees evidence in the Indus inscriptions for a Sanskrit genitive case marker:
“The genitive case-ending in Sanskrit is often -sya or -sa and in Prakrit the ending is
generally -sa or -ssa and this is what we frequently see in these inscriptions. This sug¬
gests that the Indus language is likely to have been Prakritic” (1992, 206). 48

Kak’s methodology in determining this case ending is to correlate the s phoneme in
dte BrahmT script with the most frequently occurring Indus symbol, the jar sign. He
argues that if the three types of Sanskrit s (dental, palatal, and retroflex) were counted as
one (as occurred in Pali and later prakrits where they were merged into one sibilant),
then this phoneme would occur first in the list of the ten most frequently occurring
Sanskrit phonemes, thus corresponding in frequency to the most commonly occurring
Indus sign, the ‘jar’. Most significantly, he finds that the BrahmT sign for -s does actu¬
ally resemble this jar sign. Since this sign occurs very frequently at the end of inscrip¬
tions, and since the seals of the historic period, from Ashoka’s time and after, almost
all ended in the genitive case, Kak believes he has grounds to identify it as a genitive
case marker. Seals ending in the genitive possessive marker—denoting ownership of
goods—would quite likely be used for trading purposes. Most scholars do actually be¬
lieve that the seals were connected with trade (they have been found in Bahrain and
Mesopotamia). Interestingly, Parpola (1994) agrees that this sign does occur frequently
at the end and also proposes that it could be a genitive marker (although not an Indo-
Aryan one). 49 By simply trying to identify one element of Indo-Aryan morphology in
the script, Kak can claim that the principal language of the Indus Valley Civilization
was Indo-Aryan without claiming to have deciphered the script. This is a more discreet
and effective statement, to my sensitivites, than one claiming to be able to read entire
signs. Whether it is correct, of course, is another matter.

The Indus ‘jar’ sign.

The Indus Valley Civilization 183

Dravidian Decipherment Attempts

A brief glance at how the most common Indus sign, the ‘fish,’ has been deciphered by
one or two principal scholars promoting a Dravidian language will suffice to parallel
the I'A attempts. (For a complete overview of most decipherment attempts, see Possehl
1996.) The most popular candidate for the script, from the non-Indo-Aryan side, is
Dravidian. Just as there are a priori objections against considering the script to be Indo-
Aryan, there is an a priori objection against a Dravidian identification, paralleling Lai’s
comment, noted earlier, that there is no trace of the Indus culture in the areas where
Dravidian speakers are known, and have been known, to reside. As Sjoberg (1992) puts
it: “If we assume that some Dravidian speakers at least. . . formed part of the Harappan
civilization, with its cities and its special script, the question arises: Why is there no
record of a non-Aryan writing system in South India? Writing, so far as we know, was
an Aryan introduction” (7). The same consideration applies to Munda. Moreover, as
noted earlier, the numbering system of the script, which uses a base of ten, does not
appear to correspond to that of Dravidian, which uses a base of eight.

Parpola’s (1994) work is useful exposition of everything that has been done on the
script by scholars so far with the aid of computers. 50 All these efforts are essential in
order to gain some glimpses at the possible morphological and inflectional structure of
the Indus language, but, as Parpola (1994) acknowledges, despite the significant input
from computer analysis, “we must conclude by frankly admitting our present inability
to identify morphological markers with any certainty” (97). The formal grammar of the
script being prepared by such efforts “will be useful in limiting the range of guesswork
when a breakthrough is achieved. But the breakthrough itself cannot be achieved by
this method” (101).

Anyone attempting to crack the Indus script with the presently available resources
must venture into the realm of speculation and guesswork. Parpola’s presupposition
(1994) is “that the Harappan language is most likely to have belonged to the Dravidian
family” (174). He limits his attempts at decipherment to a small number of symbols,
fully aware of the shortcomings of such endeavors. His principal decipherment involves
the common Indus sign ‘fish,’ to which he assigns the syllable min. He considers this
an appropriate identification because the word min means ‘fish’ in most Dravidian lan¬
guages and is reconstructed as such in proto-Dravidian.

The Indus ‘fish’ sign.

184 The Quest for the Origins of Vedic Culture

However, since this extremely common sign could hardly refer to ‘fish’ everywhere it
occurs, Parpola prefers to consider it a referent to ‘star’, since min is a homonym for
both ‘fish’ and ‘star’ in the Dravidian languages. He himself notes, however, that some
Harappan pottery bears symbols that specifically resemble stars, leaving one to wonder
why the symbol for another word would need to be used for the star, which has a natu-
ral-looking symbol of its own. He then finds reason to believe that die star, by extern
sion, is primarily a symbol for a god or astral divinity when it occurs in the seals. This
is quite a series of assumptions and transferrals: Parpola’s principal sign, the most com¬
mon Indus fish sign, is the symbol for a god, by association with star, which is a Dravidian
homonym for fish, which is the iconographic form that he sees in a common Indus
symbol.

Fairservis, (1992) who is also operating on the assumption that the Harappan lan¬
guage must have been Dravidian, does not believe the symbol to be a ‘fish’ at all, for a
variety of reasons, but prefers to consider it to be a loop. The Dravidian for ‘loop’ (pir)
also denotes a ‘chief or leader, which is the meaning Fairservis assigns to the sign.
Where Parpola translates ‘star’, then, Fairservis reads ‘leader’. 51 Just as there are dis¬
crepancies between Rao’s and Kak’s Indus-to-BrahmT correlations, so Fairservis differs
from Parpola in his Indus-to-Dravidian assignments, and both differ from Mahadevan,
who has also attempted to decipher die script as a Dravidian language. 52 His efforts, as
well as the dedicated and important efforts of Russian scholars, are amply outlined in
Possehl (1996) and need not detain us here. In conclusion, then, I will again refer to
Parpola (1994): “It looks unlikely that the Indus script will ever be deciphered fully,
unless radically different source material becomes available. That, however, must not
deter us from trying” (278). 53

Before concluding, we must pay heed to Witzel’s discoveries, outlined earlier, that
the linguistic influences detectable in the earliest hymns of the Rgveda are from Munda
and decidedly not from Dravidian, which surfaces only in later strata of the Rg and in
later texts. Since he holds that the Indo-Aryans entered the subcontinent at the tail end
of the Indus civilization, he infers that die language immediately prior to their arrival
was Munda—Dravidian entered later, dius influencing the later texts:

As we can no longer reckon with Dravidian influence on the early RV, . . . this means
that the language of the pre-Rgvedic Indus civilization, at least in the Panjab, was of Austro-
Asiatic nature. This means that all proposals for a decipherment of the Indus script must
start with the c. 300 Austro-Asiatic loanwords in the RV and by comparing other Munda
and Austro-Asiatic words. . . . The decipherment has been tried for the past 35 years or
so mainly on the basis of Dravidian. Yet, few Indus inscriptions have been “read” even
after all these years of concerted, computer-aided attempts, and not in a fashion that can
be verified independendy. . . . Yet, Kuiper’s ‘300 words’ could become the Rosetta stone
of the Indus script. (Witzel b, 13; italics in original)

Urbanity and the Rgveda

At dais point, we should recall that the options for the Indigenous Aryan school are
that the Vedic culture either preceded the Indus Valley culture, was contemporaneous
with it, or succeeded it. There are two other significant and commonly encountered

The Indus Valley Civilization 185

objections that would seem to controvert attempts to correlate the Vedic period with the
Indus civilization. First, smelted iron artifacts have not been found in the cities of the
Indus Valley to date, but iron is known in the Atharvaveda and the Satapatha Brahmana,
which would suggest that these texts, at least, were post-Harappan. As with the horse,
the introduction of iron has long been associated with the incoming Indo-Aryans. More
significantly, scholars from Marshall’s time to the present day have repeatedly drawn
attention to the claim that the economic landscape that can be glimpsed through the
Vedic hymns appears to be a rural pastoralist one, with no indication of urban centers.

The introduction of iron into the subcontinent by the Indo-Aryans is another dis-
carded intruding-Aryan tenet that is no longer accepted by archaeologists: “Archaeo-
logically . . . the problem of the Aryans and their association with iron remains as con¬
fusing as ever, regardless of the earlier strongly expressed theories of their apparently
tautological association” (Banerjee 1981, 320). Iron occurs in a number of locations
that could not have been influenced by one particular source. Chakrabarti (1993-94)
was one of the first to reject the idea that the Iron Age represents a “major social and
economic transformation” (25). Questioning the idea of its external origin, he claims
that iron appears in the archaeological record without causing any significant cultural
break. Here, again, he complains about “the role of the Aryans in this context which
seem[s] to have forced scholars into a position where their primary concern has been to
correlate the early Indian data on iron to some diffusionary impulses through the north¬
west” (25).

Of greater implication is Shaffer’s argument (1984a, 49) that during the late third
millennium b.c.e. iron ore was recognized and utilized in southern Afghanistan and
was manipulated to produce iron luxury items. The fact that there are early Harappan
artifacts in the same stratigraphic proveniences as the iron artifacts suggests to him that
“the ‘Early Harappan’ complexes had access to, or knowledge of, an iron technology.”
In actual fact, although there is no evidence for awareness of smelted iron technology,
iron ore and iron items have been uncovered in eight bronze age Harappan sites, some
going back to 2600 b.c.e. and earlier. (These will be described in chapter 12.) So there
was as awareness of iron, which may have been encountered accidently during the smelting
of copper, and a willingness to exploit it. The Harappan awareness of iron ore cannot
be considered an “iron age,” which is when smelted iron items became common items
of household use and occurred around 1000 b.c.e. According to Possehl (1999b), “the
iron age is more of a continuation of the past then a break with it” (165). Moreover,
iron tools did not “lead to the subjugation of indigenous population by invading war¬
riors” (164). In any event, Shaffer concludes with a complaint that has become stan¬
dard for South Asian archaeologists: “Ideas of invasions, diffusions, and conquests have
obscured and hindered investigation into the region’s indigenous cultural processes.
To fully understand and appreciate the various solutions to cultural problems recorded
in the South Asian archaeological record, alternative explanatory frameworks must be
considered” (59).

Regarding the connection between the Vedic landscape and the Harappan one, al¬
most all Indigenous Aryanists feel that the Vedic poets either preceded and/or coex¬
isted with the Harappan world. Scholars since the time of Marshall (1931) have long
since discarded the former possibility, since “if the Vedic culture antedated the Indus,
how comes it that iron and defensive armour and the horse, which are characteristic of

186 The Quest for the Origins of Vedic Culture

the former, are unknown to the latter?” (Ill). Moreover, anyone open to considering
whether the Vedic people coexisted with or authored the Indus Valley Civilization would
be “wholly at a loss to explain how the Indo-Aryans came to relapse from the city to the
village state, or how, having once evolved excellent houses of brick, they afterwards
contented themselves with inferior structures of bamboo” (112).

From the Indigenous Aryan side, there have been dozens of books attempting to fit
the hymns of the Rgveda into the ruins ol the Indus (e.g., Prakash 1966; Sankarananda
1967; Chandra 1980; Deshmukh 1982; Singh 1995). This position has to contend with
the conclusions of Wilhelm Rau, who combed the early strata of Vedic literature in
search of literary evidence for permanent Vedic settlements. His main concern was to
evaluate Wheeler’s suggestion that the fortified citadels of the Harappans must have
been those that the Vedic Aryans destroyed. Although hardly any scholars uphold
Wheeler’s position today, Rau’s remarks are very pertinent to anyone attempting to
correlate the Indus Valley Civilization with that of the Indo-Aryans, since he is evaluat¬
ing whether the Indo-Aryans had any familiarity with urban centers. His method is
primarily an analysis of the attributes of a pur , which means ‘city’ in later texts, (as noted
earlier, in the Rgveda Indra is called purandara ‘fort-destroyer’).

Rau (1976, 41) believed his work permitted him to state the following conclusions
“with confidence.” A pur in the Rgvedic period consisted of one or several concentric
ramparts on a round or oval groundplan; was built of mud or stone; was fortified with
combustible defenses; had gates made of wattle or prickly shrubs; was furnished with
wooden sheds as quarters for human occupants; was stocked with provisions for man
and cattle; was not permanently occupied but served as a refuge in times of need; was
erected in war as a base of operations; and probably needed repairs after the rainy sea¬
son. Some of Rau’s interpretations are based on inferences, which it might be useful to
reproduce so that they can be compared with those of the scholars attempting to corre¬
late the Indus cities with Vedic testimony.

That the purah consisted of concentric ramparts seems to be based on the fact that
Indra entered ninety-nine purahs in one day, on the adjective satabhujih (hundred¬
armed) which accompanies the word twice (it is also used with the river Sarasvatl),
and on the prefix pari used when the sacrificer makes a purah ‘around’ himself (Rau
1976, 24-25); that the purah were made from mud is based on the word dehi in two
verses and the emendation of vipram to vapram in a third (26); that they were com¬
bustible is based on the fact that Agni helped Indra break the purah of the enemy
(27); that they are made of wattle and palisades is based on the fact that the purah are
made to bow down, fall backward, lie on the ground, or be rent like a garment (28);
that they were fortified by prickles is “likely” based on verses produced from the very
much later Arthasastra of Kautilya (ca. fourth century b.c.e.); that the human occu¬
pants lived in sheds (vimita) is based on two verses in which a golden and mighty
vimita is said to exist in the creator Brahma’s world brahmaloke ; that the purah were
not permanently occupied seems to be predicated on verses stating that the enemy
were driven into their purah, and that the demons enter into theirs (34); that they
were erected in war, and so on, is based on verses suggesting that the gods, disadvan¬
taged without purah, decided to make some of their own; and, finally, that they needed
repairs after each rainy season is inferred from the fact that the term purah is juxta¬
posed with the word saradlh ‘autumnal’ (37).

The Indus Valley Civilization 187

There seems to be a good deal in Rau’s interpretations (1976) that is not explicit in the
texts. Nonetheless, he reiterates an often cited observation: “Not a word is said in our
texts of the characteristic features of the Indus cities, of brick walls, brick houses, brick-
paved streets laid out on an orthogonal pattern, of granaries or public baths. . . . No state¬
ment in Vedic literature prompts us to assume a . . . formidable civilization” (52).

In terms of the habitations of the Indo-Aryans themselves, Rau (1977) concludes
from the relevant terms that grama denotes “a train of herdsmen roaming about with
cattle, ox-carts and chariots in quest of fresh pastures and booty,” as well as “a tempo¬
rary camp of such a train, sometimes used for a few days only and sometimes for a few
months at the most” (203). He reads another term, said, as referring to lodgings that
“consisted of transportable sheds . . . made of bamboo-poles and reed mats, which could
be assembled and dismantled in a minimum of time” (205). Although he notes that
“there cannot be the slightest doubt that the Vedic Aryans practiced agriculture from
the very beginning,” he nonetheless maintains that this “cannot prove the existence of
‘villages’ in our sense of the term. . . . even a migrating population could do a little
tilling of the soil on the side when resting for a while” (205). Rau, then, does not allow
even villages as a feature of early Indo-Aryan society (except for some rare references),
not to mention cities or an entire civilization such as that of the Indus Valley. He finds
that the term nagara, which means ‘town’ from the time of the Taittirlya Aranyaka, occurs
only once in an earlier text. 54

In contrast to Burrow (1963), Rau considers arma or armaka to refer to rubbish heaps.
Burrow accepted Wheeler’s theory that the Aryans were responsible for the overthrow
of the Indie civilization and interpreted the terms arma/armaka as the ruined sites of
this encounter, since “during the early Aryan period, the ruins of many Indus cities
must have formed a conspicuous feature of the countryside” (Burrow 1963, 160), Rau,
in contrast, envisions herdsmen revisiting the same camping grounds by rotation and
developing the habit of throwing rubbish over the years regularly in the same spots.
This rubbish accumulated into heaps, from which only potsherds can remain in the
long run (potsherds from armas are prescribed for making certain ritual vessels). Here,
again, he concludes that the immigrating Vedic Indians by necessity must have been
constantly on the move during the centuries they conquered the plains of the Punjab
and the Ganga-Yamuna Doab. In this scenario, permanent settlements are not to be
expected at all, since the migrating trains left behind only deserted resting places. Ac¬
cordingly, “it is useless to look for structural remains dating from Vedic times, in north¬
western India” (Rau 1997, 206). If this conclusion is correct, it is puzzling since, as
Burrow noted, much of the area known to the composers of the Vedic hymns must
have been littered with hundreds of Indus sites during the time frame normally allotted
to Indo-Aryans intrusions. If the Vedic texts do not refer to such sites, it surely cannot
be because the Indo-Aryans were not aware of them. This lacuna seems to be a peculiar¬
ity of the texts rather than a reflection of the Vedic landscape. The same, as I will argue
later, holds true for any absence of references to urban centers.

Bhagavan Singh’s (1995) is a recent and perhaps the most ambitious exposition of
those holding contrary views to scholars such as Rau. Singh has compiled a glossary of
all the technical terms and items relating to material culture from the Rgveda. The ref¬
erences are extracted from all contexts—poetic analogies, metaphors, similes, and so forth—
on the grounds that the composers of the hymns could utilize such figures of speech

188 The Quest for the Origins of Vedic Culture

only if their referents were a part of their everyday environment. Rau (1976) at least
agrees with this basic point that “the world of the gods has always been fashioned in
analogy to the human environment of their worshippers”(9).

Singh’s lists are extensive. 55 I will select a few examples, which will suffice to illus¬
trate his attempt to subvert the claim that the Indo-Aryans knew no urban centers and
were simple nomad pastoralists. I should note that specialists in Vedic philology will
doubtlessly find plenty of words that Singh has translated in ways they will object to,
and even more that are fanciful to the extreme. 56 His efforts to connect the Vedic Ary¬
ans with the Indus Valley go to the extreme of suggesting that the Rbhus were seal
makers (Singh 1995, 162)! As mentioned in the introduction, I have examined evidence
from every context, since I have often found that a scholar who can make an almost
comically uncritical argument in one place can sometimes make a very insightful contri¬
bution in another. If nothing else, Singh has compiled useful lists that can provide the
basis for further careful scrutiny.

Of the seventy or so words Singh has extracted connected with cities and dwelling
places, brhantam manam sahasradvaram grham ‘very large house with a drousand doors’
and sahasrasthuna grham ‘house with a thousand pillars’ are of particular interest, since
they suggest to him acquaintance with monumental structures. 57 Likewise, from another
long list of words associated with navigation, dasaritra ‘ship with ten oars’; s'ataritra
‘ship with hundred oars’; and Vasistha in a ship in midocean make his best case. Singh
has compiled numerous words connected with government, thereby arguing the exist¬
ence of quite a sophisticated system of organization involving rastra ‘kingdoms’; a vari¬
ety of types of rulers: raja, ekaraj, samrdj, janardj, jyesthardj, and so on; and various
terms for assemblies, and similar gatherings: sarhsad, sabha, samiti, and so on. Of course,
all these concepts could have existed among nomadic people: one has no right to as¬
sume that these political terms had the same denotation in the protohistoric period as
they did in later texts. And one could always explain any genuine references to cosmo¬
politan life by postulating that the Aryans must have been aware of, and interacted with,
cities in central Asia or even the Near East. Thus Braarvig (1997), for example, explains
the reference to the hundred-oared ship—which he believes is “a type of ship scarcely
used in Vedic culture”—as an image “that had probably followed the Vedic Indo-Euro¬
pean culture from other locations where such ships were in use ... a symbol which had
got its meaning elsewhere” (347). Nonetheless, whatever their own level of social attain¬
ment might have been, there are a variety of reasons to review the long-held assumption
that the Indo-Aryans knew no urban centers. These will be touched upon in the follow¬
ing discussion and in the next chapters.

Some Western scholars have also been struck by discrepancies between the Vedic
landscape and a nomadic one. Basham (1989) comments: “It is surprising that the Aryans,
who at this time had never organized a settled kingdom or lived in a city, should have
conceived of a god like Varuna, the heavenly emperor in his glorious palace, with innu¬
merable messengers flying through the cosmos at his bidding” (12). As far as Singh is
concerned, if one is looking for nomads in the Rgveda, one will find nomads. He, at
least, does not read the texts with the same assumptions:

There was nothing in the Rgveda to look for a primitive or primarily nomadic society
[sic]. . . . Chariots and wagons and boats which occur so frequently in the Rgveda do not

The Indus Valley Civilization 189

agree well with nomadism. Movement of cars presupposes existence of roads and defined
routes, which in turn presuppose settlements and regular traffic from point to point. Boats
and ships are not floating logs. They presuppose ferry ghats and fixed destinations. . . .
there was much in the Rgveda that defied explanation. . . . But instead of reconciling the
discordant features, scholars either ignored them or distorted the facts and features. (Singh
1995, 8)

Bisht (forthcoming) has also tried his hand at establishing “points of convergence”
between the Harappans and the Rgveda. Like Rau and Singh, Bisht also pulls out ex-
tensive references from the Vedic texts that he feels could correspond to urban architec¬
tural features. He also draws attention to Mitra and Varuna in their structures contain¬
ing a thousand pillars, as well as to a settlement or house with a hundred doors, a
house with a thousand doors, a mansion with a thousand columns, a huge column
supporting tire firmament, a construction of six pillars, columns of copper, and a house
raised in the sea. A house compared to a pond particularly catches his attention: “Does
not the comparison of a house to a pond indicate that such tanks were sometimes along¬
side or within a building, the kind of which we see in the elaborate structure containing
the Great Bath at Mohenjo-Daro’’ (421), as do the pillared structures, which “remind
one of the large columnal hall in the citadel of Mohenjo-Daro” (420).

Bisht finds evidence of a tripartite settlement in a variety of terms such as tridhatu
s'arman ‘triply defended dwelling’ and tripura, which he dtinks are good candidates for
three separate divisions of an intricately fortified multi-dimensional settlement the like
of which may be seen in the Harappan city of Dholavira. . . . Harappa is also emerging
as a multi-divisioned city. . . . Kalibangan too seems to be having three divisions” (413).
This interpretation can be kept in mind when we encounter Parpola’s interpretations of
the word tripura in the next chapter. Moreover, Bisht’s purs are far more sturdy and
sophisticated than Rau’s, being drdha ‘solid’; drnhita, ‘strengthened’; adhrsui ‘impreg¬
nable’; asmanmayi ‘made of stone’; prthvi ‘extensive’; mahi ‘spacious’; ayasi ‘made of
metal, metal-strong’; and isa ‘affluent’. The satabhuji ‘hundred-armed’ forts that Rau
took to indicate concentricity are taken by Bisht as “having a hundred . . . bastions.”
Bisht makes a point, perhaps also with Wheeler in mind, of noting that tire Aryans also
owned purs; indeed, out of twenty-seven adjectives describing purs, “there occur as many
as fifteen terms that qualify the Aryan purs in contrast to only eleven or twelve in favour
of the non-Aryan ones” (410). The Aryans did not just destroy the purs of the Dasyus,
they owned their own as well.

Bisht is well aware of the plethora of opinions proffered in interpretations of the
Vedic landscape: “Diametrically opposite views have been expressed and rejected due
to intrinsic contradictions, lack of coherence or consistency inherent in the approaches,
or due to the appearance of fresh evidence” (392). Like Singh, he believes previous in¬
terpretations were predisposed to anticipating the Indo-Aryans to be “barbarian equestri¬
ans who, before entering India, roamed about in Central Asia and the Iranian Pla¬
teau . . . [and] thrived on stock-breeding and primitive farming” (392). He, too, finds
that “the information gleaned from the Rgveda projects a picture of considerably civi¬
lized Aryans.” They had “a variety of permanent settlements and fortified towns as
well as monumental structures. They were advanced in agriculture, stock-breeding,
manufacture of goods and long distance trade and commerce via roads, rivers, and
seas” (393).

190 The Quest for the Origins of Vedic Culture

B. B. Lai (1997) is a little more cautious in denying the nomadic character of tire Indo-
Aryans: “Just as there were cities, towns and villages in the Harappan ensemble (as there
are even today in any society) there were both rural and urban components in the Vedic
times. Where then is the ‘glaring disparity’ between the cultural levels of the Harappan
and Vedic societies?” (285). S. P. Gupta (1996) elaborates on this perspective:

Once it becomes reasonably clear that the Vedas do contain enough material which shows
that the authors of the hymns were fully aware of the cities, city life, long-distance over¬
seas and overland trade, etc. ... it becomes easier for us to appreciate the theory that the
Indus-Saraswati and Vedic civilizations may have been just the two complementary ele¬
ments of one and the same civilization. And this, it is important to note, is not a presup¬
position against the cattle-keeping image of the Vedic Aryans. After all, ancient civiliza¬
tions had both the components, the village and the city, and numerically villages were
many times more than the cities. In India presendy there are around 6.5 lakhs of villages
but hardly 600 towns and cities put together. . . . Plainly, if the Vedic literature reflects
primarily the village life and not the urban life, it does not at all surprise us.” (147)

Allowing the primarily nomadic life of the Vedic poets, the possibility of nomadic cul¬
tures coexisting with urban societies merits consideration. Witzel (1989) notes that the
possibility of urban centers being known to the Indo-Aryans “cannot simply be dis¬
missed,” since, although “there is no mention of towns in the Vedic texts . . . this may
be due to the cultural tendency of the Brahmins who . . . could preserve their ritual
purity better in a village than in a busy town” (245).

Potentially significant evidence against the claim that the Indo-Aryans knew no cities is
that archaeology is revealing settlements and small urban centers in the Punjab even in the
Post-Harappan period, precisely where and when few' question that the Aryans were present:

Sites such as Harappa continued to be inhabited and are still important cities today. . . .
Late and post-Harappan settlements are known from surveys in the region of Cholistan,

... the upper Ganga-Yamuna Doab,. . . and Gujarat. In die Indus Valley itself, post-Harappan
setdement patterns are obscure, except for the important sites of Pirak. . . . This may be
because the sites were along the newly-stabilized river systems and lie beneath modern vil¬
lages and towns that flourish along the same rivers. (Kenoyer 1991b, 30)

As will be discussed in the next chapter, the excavator of Pirac, situated in die Indus
Valley and dated between 1700 and 700 b.c.e. exactly where and when the Indo-Aryans
were present in the subcontinent, considers the site a town of some size with elaborate
architecture. Moreover, it revealed a more intense level of irrigation and cultivation than
existed in the third millennium b.c.e. How does this evidence fit with the pastoral,
nonurban horizens of the Indo-Aryans? The following is also noteworthy:

Although the overall socio-economic organization changed, continuities in technology,
subsistence practices, settlement organization and some regional symbols show that the
indigenous population was not displaced by hordes of Indo-Aryan speaking people. . . .

For many years, the “invasions” or “migrations” of these Indo-Aryan speaking Vedic/
Aryan tribes explained the decline of the Indus civilization and the second rise of urban¬
ization in the Ganga-Yamuna valley. . . . This was based on simplistic models of culture
change and an uncritical reading of the Vedic texts. Current evidence does not support
a pre- or proto-historic Indo-Aryan invasion of southern Asia. . . . Instead, there was an
ovedap between Late Harappan and post-Harappan communities . . . with no biological
evidence for major new populations. (Kennoyer 1991b, 30)

The Indus Valley Civilization 191

So the hymns are not necessarily sparse in urban references because of ignorance or
unfamiliarity with large settlements and urban centers. This evidence will be discussed
further in chapter 11.

Regarding the pastoral nature of the Indo-Aryans, Chakrabarti (1986) adds a further
observation that “the inconvenient references to agriculture in the Rigveda are treated
as later additions. The scholars who do this forget that effective agriculture is very old
in the subcontinent, and surely no text supposedly dating from 1500 B.C. could depict
a predominantly pastoral society anywhere in the subcontinent. Something must be wrong
with the general understanding of this text” (Chakrabarti 1986, 76). In other words, if
the Indo-Aryans were pastoralists, they must have always coexisted with agriculturists in
India since agriculture predates the assumed date for their arrival by millennia. There
could never have been a purely pastoral economic culture.

In fact, the often held assumption that the undivided Indo-Europeans, at least, were
pastoralists has been reconsidered by Indo-Europeanists. This assumption was stimu¬
lated both by the belief that the Indo-Europeans were nomadic hordes that alighted all
over Eurasia, and the fact that agricultural terms seem to be somewhat confined to the
European side of the family. Mallory (1997) points out the this belief meant that the
Indo-Iranians lost their original Indo-European agricultural vocabulary but preserved a
pastoralist one. Or it meant tht the Indo-Europeans themselves were pastoralists and
that agriculture was a later development, or encounter, by the western Indo-Europeans
after the Indo-Iranians, who preserved the old Indo-European pastoralist lifestyle, had
departed from the common home. Or it meant that the homeland contained a mixed
economy in its western parts and a primarily pastoral one in its eastern areas.

Diebold (1992), however, has reconstructed a system of settled agriculture for the
original Indo-Europeans in what he terms “Indoeuropa.” Mallory (1997) agrees that
there is solid evidence in both European and Asiatic stocks for Proto-Indo-European
cereals, as well as the agricultural terminology required to process them. He notes that
“while the economic emphasis of the immediate ancestors of the Indo-Iranians may
have been towards pastoralism there is good evidence that they too are derived from a
mixed agricultural population” (236-237) There are a variety of reasons to reconsider
the view that the character of the Indo-Aryans was purely pastoral.

While most South Asian archaeologists, in interviews and publications, seemed quite
open to the possibility that the world of die Rgveda could have coexisted with and pre¬
ceded that of Harappa, others object on archaeological grounds. In her critique of an
Indigenous Aryan book, Ratnagar (1996b) does not attempt to refute the Vedic/Harappan
correlation but does point out that the few pre-Harappan sites that have been excavated
display signs of only rudimentary technology and humble material culture: “We ask
how the fast horse-drawn chariot of the Rgveda could have been made at any of the 4th
millennium b.c.e. sites, without the metal saws, adzes and chisels that make accurate
carpentry” (75). Whereas the Rgveda is often considered too humble, materially, to be
associated with the Harappan civilization, it is here considered too sophisticated for the
Pre-Harappan culture.

This point warrants attention, although, in my opinion, one should be cautious in
attempting to connect the material from the Vedic hymns with the archaeological record.
As Elizarenkova (1992) notes, “either one comes to know things due to archaeological
findings and in this case their names and purpose may remain unknown, or only the

192 The Quest for the Origins of Vedic Culture

names of the things are known from the texts, but the things themselves, as well as
their purpose, are unknown” (129). Even if an archaeology of the Veda must be at
tempted, there are limitations in drawing far-reaching conclusions from argumentum ex
silentio in the archaeological record. Quite apart from not unearthing any saws, adzes,
and chisels for the making of chariots, we should not forget that archaeology has not
unearthed the actual chariot itself, either, until at least a full millennium after it was
known to have existed on the subcontinent. Besides, archaeology has a tendency to
suddenly unearth material that completely subverts previously held assumptions.
Mehrgarh is the prime example. Prior to its discovery, scholars were inclined to believe
that agriculture and urbanization were both diffused from West Asia. Mehrgarh, an
agricultural settlement dating back to the seventh millennium b.c.e., dramatically dem¬
onstrated that “die theoretical models used to interpret the prehistory of Southern Asia
must be completely reappraised” 0arrige and Meadow 1980, 133).

Most dramatically, Mehrgarh direw the date for evidence of agriculture back two entire
millennia. This clearly underscores the danger of establishing theories predicated on
argumentum ex silentio in the archaeological record. Mehrgarh also undermined previ¬
ous assumptions that urbanization and agriculture were diffused from the centers of
civilization to the west of the subcontinent. The site also set the stage for the indigenous
development of complex cultural patterns that culminated in the great cities of the Indus
Valley: “The origins of the Indus urban society can be traced to the socio-economic
interaction systems and setdement patterns of the indigenous village cultures of the alluvial
plain and piedmont. More importantly, the factors leading to this transformation ap¬
pear to be autochthonous and not derived from direct stimulus or diffusion from West
or Central Asia" (Kenoyer 1991b, 11).

This continuum of the archaeological record stretches from the seventh millennium
B.C.E. right down through the Early, Mature, Late, and Post-Harappan periods. Of course,
as in any cultural area over the course of time, there are regional variations and trans¬
formations, but no sudden interruptions or abrupt innovations that might alert archae¬
ologists to an intrusive ethnic group: “There were no invasions from central or western
South Asia. Rather there were several internal cultural adjustments reflecting altered
ecological, social and economic conditions affecting northwestern and north-central South
Asia” (Shaffer 1986, 230). More than everything else, this lack of cultural discontinuity
has caused an ever-increasing number of South Asian archaeologists to question: Where
are the supposedly invading Aryans in the archaeological record ? Since this invisibility is the
single most significant factor that has caused the Indigenous Aryan position to jettison
the commonly held theories of philologers and become so widespread in India in re¬
cent years, it to this that we must turn in the next chapters.

Conclusion

A primary reason that Indian archaeologists have become disillusioned with the whole
enterprise of the Indo-Aryans is because they have been offered, and initially accepted,
a progression of theories attempting to archaeologically locate the Indo-Aryans on the
grounds of the philological axiom that their nature was intrusive. These theories have
successively proved to be wrong or questionable. The course of scholarship in the last

The Indus Valley Civilization 193

century has evolved from images of blond, soma-belching, Germanic supermen “riding
dieir chariots, hooting and tooting their trumpets” as they trampled down the inferior
aboriginal Dasa (Singh 1995, 56), 58 through speakers of an Indo-Aryan language de¬
stroying the highly advanced civilization of the superior Dasa; to discrete trickles of Indo-
Aryan speakers possibly coexisting in a neighborly fashion in the cities of the Indus
Valley with the hospitable Dasa. As a result many archaeologists have become frustrated
with the whole Aryan-locating enterprise and jettisoned the linguistic claims altogether.
Failure to find any tangible evidence whatsoever of the Aryans has resulted in the present
trend among many South Asian archaeologists, which is toward considering the indig'
enousness of both the Indo-Aryans and the Dasa, period. As we saw in the greater Indo-
European problem among Western scholars, in India, too, there is a chasm between
many archaeologists and Western historical linguists, particularly since there are so few
historical linguists in India itself and so little contact with linguistic theories originating
in the West. Accordingly, the debate in India has been primarily conducted among ar¬
chaeologists, with a growing number rejecting the whole idea of anything but indig¬
enous origins for the various developments of the protohistoric archaeological record.

Vedic philology, however, being more readily at hand, cannot be jettisoned so easily
and a good number of South Asian archaeologists are quite proficient in, if not Vedic,
then at least classical Sanskrit (which may be a disqualification from the perspective of
Vedicists since words in the classical period may not have the same meanings as in the
Vedic period). Accordingly, the relationship between the Vedic Aryans and the ruins of
the Indus Valley has been negotiated in a variety of ways. Both Wilhelm Rau and Bisht,
for example, and especially Singh have taken certain liberties according to their particu¬
lar viewpoints in extracting a picture of Vedic society from the hymns and attempting to
correlate it with the archaeological record. There is little doubt that predispositions fla¬
vor the images that both have attempted to create. There is also a question of bound¬
aries. Who is deemed authorized to make interpretations, and by whom? I will only
note in this regard that Rau immediately introduces his research with a caveat that “as
a philologist, I lack the archaeological knowledge to tackle the problem in all its as¬
pects” and concludes that, since “Sir Mortimer’s theory is sustained by no literary evi¬
dence, it must rest entirely on archaeological facts and their interpretations.” He is
speaking as a Vedicist. Bisht, although clearly proficient in Sanskrit, is trained as an
archaeologist, not a philologist. He has taken it upon himself to visit the Vedic texts
as an archaeologist who has excavated extensively in the Indus Valley, particularly in
Dholavira, and is looking for descriptions of material culture that will add meaning to
the bricks and pots of his excavations. Singh is a layman. These backgrounds are rel¬
evant to issues of authority and jurisdiction in the interpretation of evidence from dif¬
fering disciplines. Must we accept all opinions as equally valid? To what extent should
scholars not specialized in a particular field be given consideration when they critique
or challenge the opinion of specialists in that area?

In conclusion, our options, then, are to locate the Indo-Aryans either before, during,
or after the Indus Valley. The references to the full-flowing Sarasvatt make the strongest
suggestion that the Indo-Aryans could have been present in the Mature Harappan pe¬
riod (although the reference to the confluence of the Sutlej and Beas must also be taken
into account). Even allowing the SarasvatT evidence, it does not, of course, prove that
they were not immigrants, albeit earlier than has been assumed, nor that they were the

194 The Quest for die Origins of Vedic Culture

dominant presence—only the script can determine that. The so-called fire altars are dubious
as indicators of an Indo-Aryan presence, but as evidence they are as strong or as weak
as anything that has been brought forward to identify diem in their overland trek through
central Asia. This should be borne in mind as the central Asian material is reviewed in
the next chapter. The lack of urban references in the Vedic texts may be a peculiarity of
these texts since settlements continued in the post-Harappan period and Pirak, at least,
was a town of some size and sophistication.

The chariot is also a dubious indicator of Indo-Aryan origins since it seems too pre¬
carious to draw overly far-reaching conclusions about such a perishable item based on
argumentum ex silentio in the archaeological record. The horse, by contrast, to my mind
remains a serious obstacle to the Indigenous position, although it has always been a
rare, imported item and also unlikely to provide much evidence in the archaeological
record. Nonetheless, the burden of proof lies with those claiming that the horse (and
chariot) were utilized in the Indus Valley, which would need to be the case if one is to
argue for a Vedic presence there. Unlike chariots, which leave little archaeological resi¬
due unless decorated with metal parts (the earliest evidence of the spoked-wheel chariot
in the Sintashta cemetery can only be inferred from soil discoloration), horse bones are
no more degradable than other animal bones, which have been found in plenty.

Of course, there are social considerations: if horses weren’t eaten, they are far less
likely to show up in settlements; and if they weren’t buried with the deceased, they would
not show up in cemeteries. Nonetheless, the horse evidence will continue to haunt the
Indigenous position. As an aside, given the insistence from the Migrationist side that
evidence of this animal must accompany any identification of the Indo-Aryans in the
archaeological record, one can be sure that if unambiguous evidence of the horse does
surface in a reliable Mature Harappan context, there will be an uproar on the subcon¬
tinent. Until such evidence is produced, however, attempts to correlate the Vedic Ary¬
ans with the ruins of the Indus Valley will have to engage in a certain amount of special
pleading.

Ultimately, the answer to the linguistic identity of the Indus Valley lies in our hands,
but it has yet to yield its secret. If the Indus script turns out to represent an Indo-Aryan
language, then I submit that most of the linguistic argument pertaining to the origins of
the Indo-Aryans, and, indeed, the proto-Indo-Europeans over the last two centuries must
be, if not completely jettisoned as some would have it, then throughly reevaluated. An
Indo-Aryan script would also suggest that these speakers were the dominant linguistic
entity in the subcontinent at this point in time and makes a far stronger case for the
possibility of Dravidian/Munda and or “language x” (y or z) intruding on an ancient
Indo-Aryan-speaking area rather than vice versa. I have attempted to thoroughly outline
the assumptions underpinning the linguistic evidence in the previous chapters, and much
of it will be instantly subverted if the script reveals an Indo-Aryan language dominating
the Northwest of the subcontinent at such an early date.

Indeed, an Indo-Aryan language during this period would have implications and
corollaries for the entire Indo-European homeland problem, if the script does indeed
go back to 3500 b.c.e., since most Indo-Europeanists hold that the Indo-Europeans
were still undivided until sometime between 4500-2500. In other words, an Indo-Aryan
script on the subcontinent at a time frame when the Indo-Europeans were still more or
less undivided would constitute a solid argument for anyone choosing to locate the Indo-

The Indus Valley Civilization 195

European homeland in India. And the assignment of a 1200 or 1500 b.c.e. date for
the Rgveda, as will be discussed in chapter 12, will merit the skepticism that Indig¬
enous Aryanists have generally directed to such efforts, as would the skepticism some
scholars have directed toward attempts at dating Proto-Indo-European.

On the other hand, if the script turns out to be any language other than Indo-Aryan,
then the Indigenous case no longer merits much further serious scholarly consideration
(although there could still have been Indo-Aryan pastoralists interacting with these ur¬
ban centers from very ancient times, even if the dominant language of the latter turned
out to be non-Indo-Aryan). But the Indigenous case, at least to my mind, will be closed.
No doubt, diehards on both sides will attempt to reconfigure things to salvage their
respective points of view if the language revealed by the script confounds their expecta¬
tions: if it turns out to be Indo-Aryan, Migrationists will likely suddenly find reason to
suppose that the migration must have taken place two millennia earlier than had previ¬
ously been thought. If it turns out to be other than Indo-Aryan, Indigenists will likely
demote their Indo-Aryans to a dominated or colonized position in the Indus Valley
(although I doubt that this is the type of secondary status for the Indo-Aryans that will
be of interest to most Indigenists). At that point both positions become even more in¬
teresting subjects for historiographical and sociological analysis in my estimation. In
particular, if the script turns out to be Indo-Aryan, and especially if the incisions on the
shards from the Ravi phase of 3500 b.c.e. are proto-Indus writings that do incorporate
the same language, the Indo-European homeland locating enterprise is likely to accrue
even more scorn from the cynics than it has hitherto—Edmund Leach’s comments at
the beginning of this chapter will take on a new significance. Whatever language is
contained in the script, in my opinion, it would be unwise for decipherers to eliminate
either Sanskrit, Dravidian, or Munda as possible candidates.

In the meantime, for as long as the script remains resistant to decipherment, possi¬
bly the path of least resistance from a philological perspective would be to suggest that
the Indo-Aryans who composed the Vedic hymns were primarily pastoralists. One has
to work rather hard to fit the landscape of the Rgveda into the ruins of the Indus Valley
civilization. This does not preclude the possibility that the Indo-Aryans might have
interacted with the urban residents of the Indus Valley, whatever language they spoke.
Of relevance here is Possehl’s observation (1977) of the “extraordinary ‘empty spaces’
between the Harappan settlement clusters,” as well as “the isolated context for a num¬
ber of individual sites” (546). He proposes that “pastoral nomads, or other highly mobile
(itinerant) occupational specialists filled in the interstices,” since such spaces are un¬
likely to have been unoccupied. He goes so far as to suggest that “pastoralists formed
the bulk of the population during Harappan times since there do not seem to be any
settled village farming communities there” (547). Pastoralists and farmers coexisted “not
... as isolated from one another, but as complementary subsystems: two aspects of an
integrated whole. One relied on the intensive exploitation of plants and arable land, the
other on the extensive exploitation of animals and pastures” (547). Moreover, “the pres¬
ence of pastoralists makes very good sense if we see them as the mobile population which
bridged the gap between settlements as the carriers of information, as the transporters of
goods, as the population through which the Harappan Civilization achieved its remark¬
able degree of integration” (548). Bridget Allchin (1977) produces case studies to demon¬
strate how “nomadic herdsmen form an important element of mral life in India and Pa-

196 The Quest for the Origins of Vedic Culture

kistan today, including the old province of Harappan culture.” She adds that “there is
every reason to suppose that they did so in Harappan times, and that they played an
important part in the economy and organization of the Harappan world” (139).

As a final note, even if the composers of the Vedic hymns did not live primarily in
the cities of the Indus Valley, this fact in and of itself does not mean that the Harappans
could not have contained Indo-Aryan speakers: a language family can obviously encom¬
pass urban dwellers as well as village dwellers, just as we see in numerous places today.
It is important to stress in this regard that anyone promoting Dravidian or Munda as the
language of the Indus Valley will anyway have to accept an identical situation: urban Dravidian
or Munda speakers coexisting with nonurban tribal Dravidian or Munda speakers (some
of which have remained tribal to this day). The southern Dravidian culture and eastern
Munda culture were radically different from that of die Indus Valley in the third millen¬
nium b.c.e., so if hypothetical urbanized Harappan Dravidians or Mundas could have
coexisted on the subcontinent with their nonurbanized fellow language speakers to the
south or east, then Indo-Aryan speakers could have done likewise. Therefore, even if
the Rgveda does not elaborately describe the flourishing cities of the Indus, there is no
way to discount the possibility that it could still have been the product of an Indo-Aryan
pastoral society that coexisted with an Indo-Aryan urban one. Moreover the Indus Valley
Civilization may very well have been multi-lingual.

As for the possibility of the Vedic landscape preceding the Indus one, here, again, if
we are prepared to overlook the horse and chariot lacuna in the archaeological record,
this is a chronological question that hinges on the date of the Veda. This will be thor¬
oughly discussed in chapter 12. For the present purposes, for all of the reasons outlined
here, it seems fair to conclude that the archaeological evidence in the Indus Valley Civi¬
lization, whatever might have been the language of its residents, has not been able to
resolve this debate. What might accomplish this is clear archaeological evidence of the
Indo-Aryan migration across central Asia and into India. So it is to this evidence that
we must now turn.

Aryans in the Archaeological Record

The Evidence outside the Subcontinent

The previous chapters have outlined the assumptions underlying the linguistic evidence
that is generally accepted as decisive in eliminating South Asia as a potential origin for
the Indo-European languages and holding that the Indo-Aryan languages must have
entered the subcontinent from the outside. The question has been raised regarding the
extent to which the various linguistic methods are capable of determining whether the
language flow of Indo-European immediately after the dissolution of the protolanguage
was from north to south or south to north, or from west to east or vice versa. And the
observation has been made that while the Vedic texts themselves can be used to dem¬
onstrate an escalating movement into the eastern and southern parts of India, they do
not provide unambiguous evidence for a movement into the Northwest itself.

The other major discipline that has been an indispensable part of the quest for the
Indo-Europeans is, of course, archaeology. While Nichols’s linguistic model could be
adopted or adapted somewhat to account for an Indo-European language spread from
an eastern point of origin, there is no archaeological evidence that can be readily in¬
voked to substantiate it. Archaeologists have not found any outgoing material culture
correlatable with the Indo-Europeans that can be traced as flowing from the east to the
west in a chronologically and geographically acceptable fashion. Accordingly, the dis¬
cussion in this section must be restricted to examining the proposals of those schol¬
ars who have attempted to use the archaeological record to trace the incoming migra¬
tions of the Indo-Aryans. This limitation is not likely to escape the attention of the
detractors of the Indigenous Aryan position: “The archaeological lack of evidence for
inward migration often cited by proponents of the ‘Out-of-India’ hypothesis would
have to be balanced with the lack of archaeological evidence for the presumably much
more massive and prolonged outward migration required under this hypothesis" (Hock
1999a 16).

197

198 The Quest for the Origin of Vedic Culture

Although it is debatable whether an outward migration would need to be much more
massive than an inward one, the point holds good. Indeed, as we have seen, Nichols’s
homeland was a priori subjected to this criticism before her thesis had even been pub¬
lished. A parallel archaeological lacuna has been one of the principal objections raised
against the central Anatolian homeland. As Gamkrelidze and Ivanov (1983a) themselves
hasten to point out in anticipation of their detractors’ reactions: “It should be noted at
the outset that the original area of distribution for die Proto-Indo-European language in
the fourth-fifth millennia B.c. does not have an archaeological culture which might be
identified in any explicit way with Proto-Indo-European” (35). As we have seen, Renfrew,
(1987) an archaeologist defending more or less adjacent geographic contours as these
linguists, does provide a material culture, but one in a significantly different temporal
bracket. However, in so doing, he completely undermines the traditional criteria used
in searching for the Indo-Europeans in the archaeological record. Since his Proto-Indo-
European is a sedentary agriculturist to be traced paleobotanically via the spread of
agriculture, he bypasses the traditionally almost exclusive focus on grave goods and pottery
styles. This raises the first very obvious and rather crucial issue: What exactly is it that
we are looking for in the archaeological record that might correspond to the speakers of
an Indo-European language? This chapter will examine this issue from the perspective
of the assumed overland routes of the Indo-Aryans on their way to the subcontinent.

Identifying Aryans

Nowadays all careful scholars begin their speculations regarding the linguistic identifi¬
cation of an archaeological culture with preambles stressing the need to be cautious
about correlating material culture with linguistic groups. Lyonnet (1994), for example,
notes that “it may be in vain to try and identify the Indo-Aryans,” since “language, eth¬
nic identity and culture are individual components that can be combined in many dif¬
ferent ways, and nothing allows us to state that, knowing language x and culture x, we
are dealing with ethnic group x” (425). At least on a theoretical level (if not always in
practice), it is by now universally acknowledged that one material culture may incorpo¬
rate more than one language group, and one language group may encompass more than
one distinct material culture. The spread of a material culture, then, need not at all
correspond to the spread of a language group, nor need a material innovation or devel¬
opment within a material culture reflect the intrusion of a new language group (although
one can also certainly not categorically deny the possibility that it may in individual
cases).

Moreover, postmodern theoretical considerations have not bypassed archaeology.
Decisions regarding how to interpret the archaeological record are no longer seen as
value-free and neutral: “It may be taken as a sign of the increasing theoretical maturity
of the discipline of archaeology that it is beginning to see itself as a product of the forces
of history. . . . there has been a shift from an internal understanding of archaeology as
an objective and value-free practice towards a broader understanding that situates ar¬
chaeology in its social and political context” (Kristiansen 1996, 139). Of particular rel¬
evance to this chapter, the adoption of migrationism as an acceptable model to account
for innovations in the archaeological record has enjoyed such a checkered history among

Archaeological Evidence outside the Subcontinent 199

archaeologists that theoreticians in the discipline have explored social contexts and
psychological paradigms to explain its rise and fall. From the times and profound influ¬
ence of preeminent archaeologists such as Kossina and Childe, innovations in the ar¬
chaeological record have typically been interpreted as evidence of some kind of a migra¬
tion or intrusion of new peoples. Allowing some variations distinguishing Continental
and American schools, this mode of interpretation remained in vogue until the 1960s,
when an antimigrational mode of archaeological interpretation started to gain the up¬
per hand. This inturn remained dominant until the late 1980s, when migrationism
again began to make something of a comeback in some quarters with appeals to not
throw out “the baby with the bath water” (Anthony 1990, 895).

The existence of such quasi-Foucaultian paradigm shifts prompted a recent publica¬
tion on the subject containing articles focusing more on the interpreters of archaeologi¬
cal data rather than on the data itself (Chapman and Hamerow 1997). Concerns fo¬
cused on “the intersection of the subjective, the inter-subjective and the objective insofar
as it relates to the use and misuse of invasions and migrations to constitute explanatory
models in 20th century prehistory” (Chapman 1997, 11). In his introduction to the
volume, Chapman muses on the formative impression that the impact (or absence) of
real-life modern migrations in the life experience of the archaeologist might have on his
or her interpretations:

There are several generations of archaeologists living in Europe whose life experiences
bore the often devastating effects of invasions and migrations in two World Wars and
their aftermaths. It is hard to resist the notion that these personal experiences did have
an effect on the models of explanation which they proposed. ... It is not a coincidence,

1 believe, that the “Retreat from Migrationism” arose precisely in countries not invaded
in either world war—in Britain, America and parts of Scandinavia, ... I suggest . . . that
the personal impact of migrations and invasions on archaeologists has been a factor much
underestimated in past “explanation” of the changing modes of archaeological explana¬
tion. I would like to suggest that there is a yet largely untapped reservoir of information
and insight about the writing of archaeological texts relating to the subjective experiences
of scholars.” (18)

Along the same lines, critics such as Champion (1990) and Megaw and Megaw (1992)
have also noted that the fortunes of migrationism can be correlated with the prevailing
political or intellectual milieu of the time. Anthony (1997), supporting the return of
migrational modes of interpretation back into “semi-respectability” in the late 1980s,
acknowledges that “the rise, fall and recovery of migration models is partly embedded
in paradigm shifts in archaeological theory, with all the socio-political factors of aca¬
demic competition that are entailed.” He notes that it is no accident that migrations
and invasions as forms of cultural change are again in vogue after two decades of ne¬
glect: “The insistent clamour of the homeless, the migrant and the refugee is rarely still
and we cannot but face its consequences on an academic as well as a human level” (21).

Nonetheless, with due precaution, it is the material culture of migrants that can pro¬
vide the only chance of physically identifying the Indo-Aryans who would otherwise
simply remain an abstract linguistic entity. Even in terms of material culture, the pros¬
pects are restricted: physical anthropology, a method upon which Aryan seekers once
pinned high hopes, is currently no longer deferred to by specialists, since skeletal re¬
mains “do not sort into ‘types’ along biological, linguistic or cultural lines.” Accord-

200 The Quest for the Origin of Vedic Culture

ingly, “the quest for the elusive Aryans lies far outside the agenda of present-day skeletal
biologists, who acknowledge the fall of the biological race concept in their discipline.
Racial palaeontology went defunct in the middle part of this century.” This forces us to
confront the crucial issue: “How could one recognize an Aryan, living or dead, when
the biological criteria for Aryanness are non-existent?” (Kennedy 1995, 61). 1 What ex¬
actly is it that we are looking for in our quest to identify Indo-Aryans in the archaeologi¬
cal record?

Our primary and only literary sources in this regard (in addition to the cultural traits
assigned to the Indo-Europeans by comparative reconstructions) are the Veda and the
Avesta. However, while Lyonnet (1994), for example, acknowledges that “both archae¬
ologists and linguists have made attempts to find (the Indo-Aryans and Iranians] in
Central Asia on the basis of both the precisions of the Rgveda and the Avesta, and ar¬
chaeological data,” she echoes the opinion of most archaeologists that none of these
attempts “are entirely satisfactory either chronologically, linguistically or archaeologically”
(425). Nonetheless, she provides a list of material characteristics corresponding to a
fairly typical version of the reconstructed material culture of the Aryans. Since this list
still seems representative of the primary features archaeologists and historians have in
mind when looking for telltale signs of the Indo-Aryans (Mallory 1989; Parpola 1994),
it seems useful to reproduce it here. According to Lyonnet (1994, 425), the literary sources
in the Rgveda provide the following information that could potentially be translated
into hard evidence retrievable from the archaeological record:

• The Aryans arc intruders, pastoralists also practicing agriculture;

• The Aryans conquer the Dasas, the local wealthy dark-skinned population who live
in or near a mountainous area, in forts that might be circular with a triple
surrounding wall; this conquest is violent, implying destructions by fire, and
involving an elite of warriors on horse-drawn chariots;

• The Aryans practice the fire cult and the sacrifice of animals (among these is the
horse sacrifice, very rare and highly prized) and even of human beings;

• They encourage cremation;

• They press soma.”

From this list, the intrusiveness of the Aryans (which is not ostensibly evidenced in the
texts, as I have noted) is the issue being debated and is not being accepted as a priori in
this discussion. The rationale behind the skin color of the Dasas has been outlined in
chapter 3 and can be legitimately ignored. Their residence in circular forts with triple
walls is an interesting interpretational proposal by Parpola, but one that will be
problematized later and need not compel us at this point.

In terms of a picture that most scholars would accept, we are therefore left with an
Indo-Aryan cultural group of pastoralists who practiced agriculture, contained an elite
warrior group, used chariots, and knew the domesticated horse. Their religion was cen¬
tered on fire sacrifices, and they practiced cremation (as well as other burial practices).
Obviously, such identification does not mean that all people who were pastoralists, or
practiced agriculture, or had a martial elite who used chariots, or even practiced fire
sacrifices were Indo-Aryans, since such characteristics were widespread in many cultures
of the ancient world at various points in history, but they are typically accepted as cir¬
cumscribing the Indo-Aryan quest somewhat. And they are the best indicators to which
we can lay reasonable claim.

Archaeological Evidence outside the Subcontinent 201

Several other assumptions need to he laid out clearly before proceeding any further.
The first is temporal and will be critiqued fully in chapter 12. For the present purposes,
I will simply outline the logic underlying the assignment of dates commonly accepted as
corresponding to the Indo-European language dispersals from the unified Proto-Indo-
European stock. I know of no objections to the terminus ante quem of Indo-Aryan as a
separate linguistic entity. The Mittani treaties have been dated to about 1500 b.c.e. (as
has the Rgveda for reasons that will be examined later). Since almost all authorities agree
that the language of the treatise is specifically Indo-Aryan, then this language was distinct
from other Indo-European languages, including Iranian, by this time at the latest.

The terminus a quo for Proto-Indo-European as a unified entity is more problem¬
atic. For now, suffice it to say that many scholars have accepted the wheeled vehicle and
related items as temporally most diagnostic (since the horse evidence has been subject
to so many objections). The rationale here, briefly, is that the cognate terms for the
wheeled vehicle indicate that this item was known to the undivided Indo-Europeans.
Since the first evidence of wheeled vehicles in the archaeological record occurs at the
turn of, and throughout, the fourth millennium b.c.e., the first assumption is that
the wheel was therefore actually invented at around this time, and the second is that the
dispersal of the various languages must have occurred after this point. The time frame
involved in the development of the Indo-Aryans as a distinct entity, then, is generally
accepted as being between 4000 and 1500 b.c.e. once other diagnostic items such as
‘plows’, ‘yoke’, ‘wool’, and ‘silver’ are factored in using the same logic oudined above.

Fiaving established a time frame of two and a half millennia or so, the next step for
most archaeologists interested in this problem is to commit to a particular geographic
route. Again, and at the risk of repetition, all of this would make much less sense with¬
out the a priori methodological imperative that “since Indo-Iranian languages are assumed
(by linguists) to have been brought into South Asia by migrants, we must begin by examin¬
ing the archaeological record for evidence of migrations, and then justify the link be¬
tween these and the spread of the Indo-Iranian languages” (Erdosy 1995a, 9; my ital¬
ics). 1 hope the preceding chapters have given the reader a clearer idea of the various
rationales underpinning these ‘linguistic assumptions’. Clearly, those who find the lin¬
guistic premises of the migration theory questionable are unlikely to be swayed by the
archaeological details that are predicated upon such foundational linguistic assumptions.
At the very least, those unconvinced by the linguistic evidence are likely to expect some
reasonably compelling degree of archaeological evidence.

We can immediately note that there is no more consensus regarding the identifica¬
tion of the Indo-Iranians in the archaeological record than regarding that of the Indo-
Europeans in general. Moreover, until relatively recent work by primarily Russian ar¬
chaeologists, the identification of the Indo-Iranians has received much less scholarly
attention than that of the western Indo-Europeans for reasons that should be obvious
from chapter 1; once it had been determined that the homeland of the Indo-Europeans
could not have stretched too far east of the Caspian Sea, the trajectory of the Indo-Iranians
was, if not superfluous, then of secondary importance to most western scholars. None¬
theless, since all scholars, whatever position they might hold on the ultimate homeland
of the Indo-Europeans, accept that the Indo-Aryans, at least, entered India from the
West, almost all who have attempted to trace the itinerary of the Indo-Aryans do so by
tracing the latter’s itinerary from the vicinity of some part of the Caspian Sea.

202 The Quest for the Origin of Vedic Culture

There are two obvious routes to the Indian subcontinent from the Caspian Sea area:
a northern route from the northeast of the Caspian Sea through the steppes of central
Asia and down through Afghanistan and into India; and a southern route from the
southeast of the Caspian Sea through die deserts and plains of northern Iran and into
Afghanistan. The two routes have diametrically differing archaeological cultures. Both
have been identified with the Indo-Aryans in a variety of ways by different scholars, and
both have received extensive criticisms from promoters of different routes. The prin¬
cipal material involved will be examined in the following sections. I request the
nonarchaeologists to bear with the sometimes tedious details of relevant archaeologi¬
cal minutiae, which have been minimized and simplified as far as possible; even a
minimally comprehensive study of the Indo-Aryans must unavoidably plow through at
least the basic archaeological material at stake.

The Northern Route
Vedic Burial and Funerary Practices

The northern route is based primarily on the evidence of grave sites. These graves con¬
tain traces of material culture that are perceived as corresponding to items of Indo-Aryan
culture as understood from Vedic texts. The first task, dren, is to glean any information
pertaining to burials drat is available in these texts. Rgveda 7.89 speaks of going to the
‘house of clay’ mrnmdyam grham (which has been correlated with the kurgan burials);
10.18.10-13 also clearly speaks of burial in the earth. However, some scholars see signs
of cremation a few hymns earlier (10.16.1-6) were Agni is requested not to fully con¬
sume the body but to send the deceased to the forefathers after he had been fully cooked.

The Satapatha Brahmana (8.1.1 ff.) gives a variety of recomendations pertaining to
the appropriate geographic location for burial (such as near waters dowing in a south¬
easterly direction). It prescribes the construction of tombs with four corners and the
size of a man (leaving no room for another). For a Kshatriya, the sepulchral mound
should be as high as a man with arms upstretched; for a Brahmana, it should reach up
to the mouth; for a woman, up to the hips; for a Vaisya, up to the hips; and for a
Shudra, up to the knee (having said all this, the text then states that one should radrer
make the tomb so that it reaches below the knee, underlining die previous injunction
that this would thereby leave no room for another—multiple burials seem to be clearly
discouraged in this text). The text continues to say that some bank up the site after
covering the mound. It is then enclosed by means of stones. Datta (1936), who has
compiled all the references to burial practices in the earliest Vedic texts, notes that, in
contrast to this description, the later Asvalayana Grhya Sutra describes how the body is
burned, and the bones subsequently gathered and placed into lid-covered urns (with
special markings for men but not for women), so cremation and urn burials seem to be
practiced in this later period.

Since the Agni hymn just mentioned is generally considered to be a later hymn, it is
possible that burial was the older practice, which continued in later times along with
crematory practices (although one must be wary of basing far-reaching conclusions on
argumenti ex silentio ). One could therefore argue that burial practices characterized the
Indo-Iranians and Indo-Aryans in the earlier part of their trajectory, and that the prac-

Archaeological Evidence outside the Subcontinent 203

tice of cremation developed as they neared the subcontinent. However, as Erdosy notes,
cremations are rare in central Asia (being evidenced only in the cemeteries of southern
Tajikistan), and, “on present evidence, cremations appear to have originated in the Indo-
Iranian borderlands and spread northwest (and southeast) thence, against the postu¬
lated movements of Indo-Aryan speakers” (11). In short, Vedic burial practices leave us
with too many possible options—burials, cremation, and postcremation urn burials—to
be of much real use in identifying specific Aryan graves in the archaeological record to
the exclusion of others.

The Pit Grave (Jamna) Culture and Related Cultures

I have already outlined how the Kurgan culture is particularly favored by some scholars,
since the horse has been domesticated there since at least the fourth millennium b.c.e.
by people practicing a pastoral economy. Wheeled carts are also known there since the
third millennium B.C.E. I have noted Talageri’s (and the Russian linguists’) calling into
question the assumption that the place of domestication of the horse must correspond
to the original homeland of the Proto-Indo-Europeans; from their perspective, the horse
could have been known to the undivided Proto-Indo-Europeans without they themselves
being inhabitants of the locus of domestication, or personally being the domesticators.
There has been a litany of objections to the Kurgan homeland theory, which have been
touched upon throughout the previous chapters, but, as will be obvious by now, Indig¬
enous Aryanists are primarily interested in the history of the I ndo-Aryans; from their side,
there have been very few critiques of the various homeland proposals of the pre-Indo-
Aryan, undivided Proto-Indo-Europeans.

Talageri, an exception in this regard, has examined Gimbutas’s Kurgan theory and
rejected it using arguments similar to those of some of the scholars noted in chapter 1.
He finds Gimbutas’s assignment of Indo-European attributes to archaeological artifacts—
such as construing a figure with a mace in his hand as a male thunder god divinity—far
too generalized and vague to be meaningful. He also objects that the items Gimbutas
has promoted as reconstructable Proto-Indo-European—the use of copper, vehicles, boats,
and so on—are actually shared by many Old World cultures that were not Indo-Euro-
pean-speaking. He argues that the same could be said of some of the social and reli¬
gious features brought forward—male dominance, sun worship, and so forth. For the
most part, however, Indigenous Aryanists are not particularly impressed or concerned
with the homeland speculations of their Western colleagues or with the trajectories of
the other Indo-European languages. It is the Indo-Aryans who are of concern to them,
and it is with the evidence concerning this side of the language family that they are
prepared to debate.

In somewhat parallel fashion, apart from relatively recent work by Russian archae¬
ologists and one or two other notable exceptions (e.g., Parpola), Western scholars have
traditionally paid very little attention to the archaeological traces of the Indo-Aryans,
focusing primarily, and sometimes exclusively, on the origins of European civilization.
In a recent edited collection of her articles, Gimbutas, who has otherwise written pro-
lifically on the Indo-Europeans, makes barely a passing reference to the Indo-Aryans.
Her only comment is that “the point of origin for the Old Indie people ... [is most
probably] migrations of the so-called Tazabag’jab Bronze Age culture, kin to the proto-

204 The Quest for the Origin of Vedic Culture

Scythian Andronovo and timber-grave culture of the Eurasian Steppes . . . around the
15-14th centuries bc, the date which agrees with the destruction of the walled cities of
the pre-Aryan civilization” (Gimbutas 1997, 14). Mallory (1975) notes that “archaeo¬
logical solutions to the IE problem. . .have involved considerable concern with the cul¬
tural development of central and eastern Europe, possibly at the expense of analyzing
the problem in terms of also explaining the Indo-Iranian migrations where archaeologi¬
cal evidence is still quite scarce” (344). He does, however, dedicate an adequate section
to this group in his 1989 book and in a recent article, which will be discussed later.
Renfrew (1987), was torn between two hypotheses for India that are diametrically op¬
posed both temporally and culturally (A: a Neolithic agricultural incursion, which would
make the Indus Valley Indo-Aryan; and B: the traditional Post-Harappan, Steppe inva¬
sion model); this simply underscores the problematic nature of identifying this language
group in the archaeological record. (Since 1988, Renfrew has been in favor of the first
of these options.)

The northern route typically commences with the Kurgan culture. This culture was
initially identified by Childe (1926), who believed that the south Russian steppes “cor¬
respond admirably to the character of the Aryan cradle as deduced by linguistic
palaeontology. . . . The remains in question are derived almost exclusively from graves
containing contracted skeletons . . . surmounted by a mound or kurgan” (183). More
recently, Gimbutas has been the principal advocate of this position. This view has re¬
ceived plenty of criticisms, which were outlined in chapter 1 and can be encountered in
the works of anyone promoting a different homeland. They need not detain us here,
since our concern is specifically the Indo-Aryans. The Kurgan (Pit Grave) culture in the
Pontic-Caspian steppe (3500-2800 b.c.e.) evolved into the Hut Grave culture (2800—

The northern route.

Archaeological Evidence outside the Subcontinent 2 05

2000 b.c.e.), which in turn was succeeded by the Timber Grave (Srubnaya) culture (2000-
800 b.c.e.) and the related Andronovo culture (1800-900 b.c.e.), which covered an
enormous area from south of the Urals, across Kazakhstan, and into southern Siberia.
(For a comprehensive classification and periodization of Androvo Cultural Sites see
Kuzmina 1985.) The Andronovo culture is especially associated with the Iranians (or
sometimes Indo-Iranians or Indo-Aryans) by Soviet archaeologists. 2 The identification
is fortified by the fact that these cultures were the direct ancestors of the Iranian-speaking
Scythians and Sakas in historical times and fit appropriately with Lyonnet’s criteria. 3 As
is evident from the names of these steppe cultures, graves are used to inter the dead,
and the various cultures are to a great extent characterized by distinctive burial arrange¬
ments. 4 A variety of pottery types found in the graves is also of relevance.

Kuzmina (1994) narrows her criteria for identifying the Indo-Iranians down to one
positive item (that of horse-drawn chariots) and two negative ones (they did not know
temples, nor [following Rau 19741 the potter’s wheel). This last criterion requires that
“one has to look for a culture with hand-made pottery” when attempting to locate the
original Indo-Iranians (404). As for chariots, Kuzmina argues that the grave goods from
the cemeteries of the Sintashta-Petrovka culture in the southern Urals—of which the
Sintashta cemetery is most notable—and in northern and central Kazakhstan show
the earliest remains of this technology. As I have discussed, a chariot with spoked
wheels dated to about 1 700-1500 b.c.e. was found in the Sintashta cemetery, an early
Andronovo burial in the Kazakh steppes of the southern Urals. This, coupled with a
variety of weapons also found as grave goods, are associated with paraphernalia utilized
by the Indo-Aryans (as evidenced by the Hurrian Mitanni texts on horsemanship and
by the Rgveda).

Anthony (1995a) notes that horses were often sacrificed in the mortuary rites of the
Sintashta culture, which he attempts to correlate with a hymn from the Rgveda wherein
a horse is offered to the gods. He especially draws attention to one burial that contained
the corpse of a decapitated victim whose head had been replaced by that of a horse. He
finds reason to connect the fate of this individual with the Dadhyanc myth in the Rgveda
(1.116). In brief, Dadhyanc was given a horse’s head through which he could tell the
Asvins about Soma and the beheading of the sacrifice. This horse’s head was then cut
off by Indra and replaced with his own head. Aldiough, the context of this myth has
nothing to do with burials or funeral rites, the attempt to correlate this story with the
contents of a solitary grave does gives some indication of the paucity of evidence avail¬
able to archaeologists in the quest for the Indo-Aryans. In any event, Anthony (1998)
suggests that the instance of this burial custom in the Andronov culture “might there¬
fore be seen as something more significant than just the spread of a new burial custom.
It might well represent the adoption of a larger Indo-Iranian identity, a necessary part of
which was the Indo-Iranian language” (108).

One constantly finds archaeologists interpreting reliefs on seals or pottery, or even
innocuous items of material culture, in terms of Vedic or Avestan religious narrative.
Gening (1979), for example, also underscores the Indo-Iranian identification of the
Sintashta graves, but, takes major liberties in assigning Vedic significance to the various
features connected with the graves. As far as he is concerned, the fact that the horse
skeletons found in the cemetery were always unharnessed reflects the unharnessed
wanderings of the horse before the asvamedha-, the dog burials are related to motifs from

206 The Quest for the Origin of Vedic Culture

the much later Mahabharata and to the dog Sarama, which helps Indra find the cattle;
the remains of straw and posts covering the grave conjure up the barhis, or sacred straw
covering the sacrificial spot. Klejn argues against an Indo-Iranian or Indo-Aryan identi¬
fication of the Andronov culture, but some of his correlations also display a question¬
able degree of interpretative liberty: the red-ocher powder covering the skeletons in the
catacomb graves is connected with the red powder used in modern Indian weddings;
the stone battle-ax found in many graves, with Indra’s vajra; stone pestle grave goods,
with soma-pressing implements; the concave walls of the graves, with the Vedic vedi
sacrificial altar. Clearly, if one is so inclined, any innocuous grave detail can be con¬
nected with something from the gamut of Sanskrit literature and interpreted as proof
of the Indo-Aryans.

According to Kuzmina, the fact that the essential equipment of the Indo-Aryan chari¬
oteers in the Mitanni kingdom and in India has no prototypes or analogies in either the
Near East or Harappan India, but rather does show affinity with the items in the Sintashta-
Petrovka burials mentioned earlier, “corroborates the hypothesis that locates the Indo-
Iranian homeland on the Eurasian steppes between the Don and Kazakhstan in the 16th—
17th centuries bc.” She adds, appropriately, that “to dispel all doubts we have only to
find warrior burials similar to those of the steppes in Mitanni and in the northern parts
of the Indian subcontinent” (Kuzmina 1994, 410). These have yet to be found. Where
Kuzmina finds Andronovo archaeological prototypes for the inferred Indo-Aryan cul¬
tural equipment known by the Mitanni Syria in the Near East and the Vedic speakers
in India, Klejn points out that no actual trace of dais Andronovo culture in the archae¬
ology of either of these-Indo-Aryan cultures in the Near East or India has come to light.
Klejn’s critique of this Andronovo hypothesis raises important objections. While ac¬
knowledging the Iranian identification of the Andronovo culture, he finds it much too
late for an Indo-Aryan identification, since the Andronovo culture “took shape in the
16th or 17th century b.c, whereas the Aryans already appeared in the Near East not
later than the 1 5th to 16th century B.c.” More important, “these [latter] regions contain
nothing reminiscent of Timber-Frame Andronovo materials” (Klejn 1974, 58). This is
an essential point, especially since, as we have seen, some scholars date the Indo-Aryan
presence in the Near East to the 18th or 17th century b.c.e. How, then, could the Indo-
Aryans have been represented in a completely different material culture in the steppes
at more or less the same time? An Indo-Iranian affiliation of the graves is even more
unrealistic, since the joint Indo-Iranian period would have been much earlier than the
dates for the Andronovo period. Brentjes (1981), we can recall, pointed out the same
objections with the Andronovo theory.

As for India, as Lyonnet (1993) notes: “To this day no traces of such stock breeders
have been detected south of the Hindukush” (82). This is the most serious, and obvi¬
ous, shortcoming of the Andronovo Indo-Aryan or Iranian identification. Francfort (1989)
stresses this point: “Nothing allows us to dismiss the possibility that the Andronovians
of Tazabagjab are the Indo-Iranians as much as the fact that they vanish on the fringes
of sedentary Central Asia and do not appear as the ephemeral invaders of India at the
feet of the Hindu Kush” (453). A later Iranian affiliation of the Andronovo culture is
sometimes suggested, although, even here, Bosch-Gimpera (1973) objects that “there is
nothing in Iran in the second millennium that is related to Andronovo, something which
one would expect if the cradle of the Indo-Iranians were to be found in this territory”

Archaeological Evidence outside the Subcontinent 207

(515). Such Archaeologists of the region are quite specific that “the notion of nomads
from the north as the original Iranians is unsupported by the detailed archaeological
sequence available” (Hiebert 1998, 153). As far as Sarianidi (1993b) is concerned, the
Andronovo tribes “penetrated to a minimum extent. . . not exceeding the limits of normal
contacts so natural for tribes with different economical structures, living in the border¬
lands of steppes and agricultural oases” (17).’

Beshkent and Vakhsh Culture

The Beshkent and Vakhsh culture, in southern Tajikistan, known so far only by its
cemeteries, has also been related to the Indo-lranians: “In the Beshkent cemeteries we
have cremation rites; ritual hearths were built in the graves; and swastikas were used in
marking the site. In the Vakhsh cemeteries funeral pyres were lit around the grave of a
leader. A number of beliefs and cult practices that can be reconstructed from the mate¬
rials found in the Vakhsh cemeteries recall common Indo-European rites and beliefs or
specifically Indo-Iranian ones” (Litvinsky et al. 394). With these hearths we again find
hasty correlations with later Vedic characteristics: “The sacred fires of India hold the
key. . . . the existence of round and square hearths-altars in ancient India ... is identi¬
cal with the phenomenon we find at the sites in the West Pamirs” (Babayev 1989, 93).
Yet the Indian hearths being referred to are the garhapatya and daksina fires, which are
performed in sacrificial contexts that have nothing to do with burials.

The primary characteristics of the cemeteries is that they are of the kurgan type, typi¬
cal of the steppe people. The pottery, on the other hand, is comparable with the ware
produced by the sedentary agriculturists of northern Afghanistan. Litvinsky and P’yankova
interpret these two influences as a fusion of cultures between southern sedentary agri¬
culturists from Bactria and Indo-Iranian steppe pastoralists from an Andronovo proto¬
type. They find reason to consider them to be proto-Iranians (following an earlier Indo-
Aryan wave) sweeping eastward from sites like Namazga VI in southeastern Turkmenistan,
who pushed up into the valleys of southern Tajikistan, where they adopted burial rites
and certain economic systems from die more northern Andronovo steppe cultures. (See
also Piankova 1982 for an analysis of the Tajikistan culture.) Other scholars, such as
Vinogradova (1993), primarily stress the Andronovo proto-forms for the graves. The
original excavator of the site, Mandel’shtam, and more recent scholars such as Klejn
(1984), have considered these graves to be of Indo-Aryans but unconnected to the
Andronovo culture.

Like all other archaeological cultures, there is no unanimity concerning the Indo-
Iranian or Indo-Aryan ethnic identification of these burials either. Lyonnet wonders
why, if they had been Indo-Aryans who had provoked or appeared at the time of the
collapse of the Bactria and Margiana Archaeological Complex (BMAC), Namazga, and
Harappan civilizations, they did not continue to foster the links between these regions,
which had previously been connected for millennia. Rather, these connections collapse
at this time (Lyonnet 1993, 83). She underscores the extreme paucity of metal objects
found in the graves, which “is rather odd for a culture considered to come from the
Andronovo people, famous for their metallurgy” (Lyonnet 1994a, 430). Moreover, “no
trace of the horse is found, there is no evidence of any social differentiation, and, alto¬
gether, the material is rather poor” (430). As far as she is concerned, “if we are dealing

208 The Quest for the Origin of Vedic Culture

with intruders, as some features suggest, and if it is certain that they are not Andronovians,
we do not have enough evidence to identify them as Indo-Aryans. We can only com¬
pare their movement to the textually known much later migrations of two other groups,
who, coming from ‘the steppes,’ went through Central Asia into India.” These are the
Kusanas around the beginning of our era, and the White Huns in the fifth century
A.D.: “All these nomads, albeit at different periods, took exactly the same path, used
exactly the same areas for their cemeteries consisting of kurgans that all look alike from
the outside” (430).

Tribes that bury their dead in kurgans (which are so common over vast geographic
and temporal expanses) have been migrating into India throughout its history, but
these have not induced language shift across the entire north of the subcontinent. So
one is hardly compelled to interpret the scanty evidence of the Bishkent and Vakhsh
cultures as evidence of the arrival of a new language group on its way to Indo-Aryanize
North India. Like the Andronovo culture, this culture does not enter the subconti¬
nent either. Moreover, Piankova (1982) dates the graves to the last quarter of the second
millennium, which is far too late for migrants who are supposed to already have com¬
pletely settled down and written the hymns in the Indian subcontinent by this time,
even allowing the lowest possible dates proposed by scholars for the Rgveda. More¬
over, anyone prepared to gloss over the absence of horse bones in these sites cannot
then deny the presence of the Indo-Aryans in the Indus Valley Civilization on these
particular grounds.

The Southern Route

On the grounds of such objections, the correlation of the Indo-Aryans with the Andronovo
culture has been rejected by a number of scholars who prefer to opt for a southern route.
Bosch-Gimpera (1973) notes that “it would seem more likely that we should seek for the
antecedents of the Indo-Iranians by tracing their subsequent migration into Iran by way
of the Caucasus, not across the Oxus” (51 5). 6 He interprets various artifacts found in the
Indus Valley (such as the copper weapons and other items that will be discussed later) as
indicative of “an immigration of people into India with a culture related to northern
Anatolia, the Caucasus and possibly even to Mesopotamia” (517).

This southern route, across the Gorgan Plain in southern Turkmenistan and north¬
eastern Iran, is littered by the sites of sophisticated urban centers. It has long been rec¬
ognized that many of these sites, such as Tepe Hissar III, seem to have been abandoned
roughly at the same time after a period of urban florescence. 7 According to Kohl (1984),
it is debatable “whether or not the collapses were due to the cessation of overland long¬
distance trade or, say, to the incursions of steppe nomads from the north or to some
other unspecified cause” (226-227). The possibility of an invasion of these sites by Indo-
Iranian steppe nomads from the northern steppe route was previously accepted by most
scholars and is still entertained by some: “A widespread destruction of the north Ira¬
nian sites such as Tepe Hissar I1IB coupled with the equally widespread lack of ar¬
chaeological evidence of the early second millennium B.c. . . . are often associated with
peoples variously designated as the Indo-Iranians, Aryans or proto-Indic Peoples” (Tho¬
mas 1992, 20; see also Dyson 1968; Masson 364b 1992). However, this interpretation

Archaeological Evidence outside the Subcontinent 209

The southern route.

is falling into disfavor among a number of archaeologists of this area, since “there is no
documentary evidence of steppe cultures invading agricultural oases. . . . Archaeological
data was found but failed to convince. More or less intensive contacts were an established
fact, but . . . archaeological diggings testily to the fact that various influences, including
mass migrations, moved only from south-west to north-east” (Khlopin 1989, 75-76). 8
Hlopina (1972) noted that “tire excavations of the Bronze Age sites show no trace of de¬
struction,” and that “it may well be that... we shall have to abandon the hypothesis that,
at the end of the 2nd millennium before our era, the agricultural sites in these areas were
subjected to the influence of the livestock-raising tribes of the steppes and their cultures,
and that this is why these sites entered on a period of tangible decadence” (213).

A substantial body of scholars consider these urban cultures themselves to have been
Indo-Iranian in some form or fashion, rather than being invaded by Indo-Iranian steppe
nomads from the north. Roman Ghirshman (1977) has been a particularly influential
proponent of the Indo-Aryan nature of the southern route. As far as he is concerned,
the culture of the Andronovo tribes has no semblance whatsoever with that of the Indo-
Aryans. In contrast, Ghirshman builds up a case on the Grayware found in sites such
as Tepe Hissar III in the Gorgan Plain of northern Iran, which “is different from every¬
thing we know in Iran, Mesopotamia and Asia Minor” (12). But he does find connec¬
tions between this and the innovative black ceramics of the Mitanni kingdom (8). The
discovery of a seal in Tepe Hissar III containing a motif of a horse-drawn chariot and
of possibly horse-controlling trumpets reinforces this region’s connection with the horse-
and-chariot-using Mitanni Indo-Aryans. 9 All this, for Ghirshman, points to the arrival
of Indo-Aryans in the Iranian plain of Gorgan.

In this version of events, the Indo-Aryans of the Gorgan Valley, under pressure from
nomadic invasions from the north (by the very people whom some other archaeologists

210 The Quest for the Origin of Vedic Culture

consider to be the same Indo-Aryans!), fled their homes. Ghirshman’s evidence for this
assertion is the identification of houses burned by fire and the existence of arrowheads.
One group headed south toward Syria to become the Mitanni of Syria. Another branch
proceeded farther east; evidence of their journey is inferred from the appearance of black
pottery along the Kopet Dagh chain in sites such as Namazga IV, where it appears in
distinction to the pottery preceding it. 10 In time, the Indo-Aryans were expelled from
Namazga IV, too, by the same nomadic steppe tribes, and proceeded east. Ghirshman
(1977) tries to find further evidence of the Indo-Aryans in the Balkh region of Afghani¬
stan but is forced to account for the different, clear-colored pottery found there as “in¬
dicating a change of taste” (40).

Scholars arguing for an Indo-Aryan element in southern sites take the same interpre¬
tative liberties as those interpreting grave goods in the northern steppes to support their
case. Kurochkin, for example, sees an Aryan presence in a pestle and mortar wirh a
spout found in the royal cemetery of Marlik in northern Iran, which he compares with
a Siva linga. It should go without saying that upholders of an Aryan migration theory
generally consider the Siva lihgam to be a pre-Indo-Aryan icon from India, which would
have no connection with any hypothetical Indo-Aryans in northern Iran.

As with the northern route, there is no dearth of criticisms that have been levied
against such a southern route. 11 One objection points out that “the similarity is not
very close” between the Mitanni pottery and that of the Gorgan plain (Parpola 1994,
148). Deshayes (1969) notes that “the grey ware did not at that date spread beyond the
limits of the plain of Gorgan” (13). Others object that even within the spread of this
ware, the Mitanni were not the only language group present; there were non-Indo-Euro-
pean language groups such as the Lullubu, Guti, and Hurrians (Kuzmina 1993, 403).
Grantovsky (1981) complains that the gtayware can be traced back to the fourth millen¬
nium b.c.e. and is therefore too early for the Indo-Iranians. He also notes that the Indo-
Aryans are considered to be pastoralists and therefore cannot be connected with the
southern urban sites, with their agricultural base. Kuzmina (1981) draws attention to
the lack of horse bones in southern central Asia and Iran, which corresponds poorly
with horse-using Indo-Aryans.

Cleuziou expresses frustration that the appearance of the grayware is typically accepted
as heralding Indo-European invaders (with some scholars promoting an eastern move¬
ment, others a western movement, and still others a movement in both directions), and
then its disappearance is also considered to be due to Indo-European invaders! He ech¬
oes objections to such reconstructions that are by now quite standard: “Migrationists
consider that movement of people is responsible for the movement of pottery assem¬
blages, and they think that it suffices to demonstrate that potteries have moved to dem¬
onstrate the migrations” (Cleziou 1986, 244). He notes that no one has yet come up
with an archaeological origin for these hypothetical Grayware people. This “remains a
challenge for the advocates of the migratory hypothesis” and prohibits confirmation of
“the hypothetical reconstructions of philologists” (232-233). As far as he is concerned,
“No migration from outside is necessary to explain its [the grayware’sj development in
the Southeast Caspian area, and since it is the necessity of such a migration which is the
ultimate argument of its advocates, this can no more be regarded as conclusive” (236;
italics in original).

Archaeological Evidence outside the Subcontinent 211

Here, Cleuziou has articulated die main point that 1 am attempting to underscore in
this chapter. It is the verdict of linguistics that is compelling archaeologists to interpret
practically every innovation in the archaeological record between the Caspian Sea and
India as possible evidence of Indo-Iranians. As far as Cleuziou is concerned, the spread
of Indo-European “is a philological concept and we entrust the philologists to discuss
about its relevance” (245). As far as the archaeology of the southeastern Caspian area is
concerned, one could just as well argue that “local evolution led to a regression of ur¬
ban centered settlements and of settled agriculture, probably with an extension of pas¬
toral economy, and the early Iron Age represents a renewal of settled agriculture with
re-occupation of some areas.” There is no need to postulate the intrusion of foreign
tribes: “Whoever were the people concerned is not of interest here, as archaeology will
probably never tell it” (244). Moreover, and equally relevant, it is important to again
stress that “interaction between nomads and sedentaries does not imply different people
(nor different languages)” (247). I will return to this shortly.

Those who do find reason to connect the trajectories of pottery with that of the Indo-
Aryans need to address one line of argument that will disqualify the Indo-Aryans from
having any connection with wheel-made pottery at all. Wilhelm Rau has compiled the
Vedic references to pottery from the oldest strands of the Black Yajurveda and found
that although the potter’s wheel was known, it was hand made pottery that was pre¬
scribed for the ritual sphere. This suggests to him that “the more primitive technique
persisted in the ritual sphere while in secular life more advanced methods of potting
had already been adopted.” Should this assumption be correct, “we can pin down the
transition from hand-made to wheel-thrown pottery, as far as the Aryans are concerned,
(down) to the earlier phases of Vedic times” (Rau 1974, 141 ). 12 Of relevance to this line
of argument is a verse from the Taittirlya Sarhhita (4, 5, 4), stating that what is turned
on the wheel is A suric and what is made without the wheel is godly (e.g., Kuzmina
1983, 21). According to Rau’s philological investigations, the characteristic of this old¬
est pottery was that it was made of clay mixed with various materials, some of them
organic, resulting in porous pots. These pots were poorly-fired and ranged in size from
about 0.24 m to 1.0 m in diameter at the opening and from 0.24 m to 0.40 m in height.
Furthermore, they showed a lack of plastic decoration and were unpainted (Rau 1974,
142). Of further relevance is the fact that firing was accomplished by the covered baking
method between two layers of raw bricks in a simple open pit. In later times this was
done with materials producing red color. Rau advises excavators to be “on the lookout
for ceramics of this description among their finds” (142).

Kuzmina (1983), at least, has taken this advice seriously. As far as she is concerned,
“all . . . evidence as to the character of the pottery produced in Asia Minor and Central
Asia in the third and second millennium b.c. categorically rules out searching for the
proto-home of the Vedic Aryans throughout [this] entire stretch” (23). According to her,
then, the southern route is ruled out. In contrast, she holds that on many essential
points Andronovo pottery techniques are absolutely similar to those practiced by the
Vedic Aryans (as reconstructed by Rau): “Ceramic finds trace the gradual infiltration of
the farming oases of Marghiana and Bactria by the late-Andronovo tribes and their
emergence on the mountain passes leading into the Indian subcontinent, which may
provide the clue to the problem of the origin of the Aryans” (24-27). Kuzmina is forced

212 The Quest for the Origin of Vedic Culture

to concede, however, that “in the Andronovo culture it was mainly the womenfolk who
engaged in the making of pottery. ... in the case of the Vedic Aryans it was the male
paterfamilias.” 13 Moreover, “The second major distinction is the richness of the im¬
pressed decoration of the Andronovo pottery, whose geometrical designs include tri¬
angle, meander, swastika, lozenge and herringbone” (26). Vedic pottery is supposed to
be plain. Neither southern nor northern routes, then, have fully fulfilled Rau's Vedic
pottery criteria.

Sarianidi (1993c), another prominent adherent to a southern route, agrees with
Ghirshman on the basic thesis that the Aryan tribes “should not be derived from the
pastoral tribes of the Andronovo culture of the vast Asian steppes but—on the contrary—
from the highly developed Near Eastern centres” (256). While disagreeing with Ghirshman
about the grayware pottery/Indo-Aryan correlation, Sarianidi accepts the difference between
Hissar III and previous levels at that site. He further notes the commonalties between
Hissar III and the BMAC, which he considers evidence of the presence of Indo-Irani-
ans migrating into northern Afghanistan from Iran. Sarianidi is sympathetic to an ulti¬
mate Near Eastern Proto-Indo-European homeland as argued by Gamkrelidze and Ivanov.
As far as he (1997) is concerned, “it may be considered proved that the peopling of the
Bactrian Plain was the result of the arrival of tribes from the West” (644). 14 However,
this westerly origin for the BMAC, previously accepted by scholars (i.e., e.g. Kohl 1981),
is presently under reconsideration by scholars such as Hiebert (1995, 200).

The Bactria and Margiana Archaeological Complex

The BMAC culture, which has been dated to the turn of the third to second millen¬
nium b.c.e ., 15 was excavated initially by Sarianidi and his colleagues and more recently
by Fredrik Hiebert. Sarianidi (1993a) considers the whole BMAC complex to have been
specifically Indo-Atyan partly due to the discovery there of seals bearing religious motifs
similar to those found in northern Syria during the time of the Indo-Aryan Mitanni
kingdom. This identification is reinforced by other objects common to the two areas—
miniature stone piles and ritual vessels—and the discovery of axes with horse heads found
in some of the BMAC graves.

More dramatically, perhaps, in the BMAC site called Dashly-3, Sarianidi (1977) found
ash pits raised on brick platforms in a circular “temple." Also reported from Dashly-3
is “the occurrence of a shrine inside the fortress with an altar against the wall [which]
validates a suggestion that this was a ceremonial centre, probably a temple with numer¬
ous services, repositories, granaries, dwelling-houses for priests and auxiliary person¬
nel” (Tosi et al, Shahmirzadi and Joyenda 1992 220). In addition, in the southeastern
Kara Kum desert, other sites on the western limits of the BMAC culture that have been
labeled-“temple-forts” (Parpola 1995) or “ceremonial centres” (Jettmar 1981) have been
found. Most striking of these are the sites of Togolok-21 and Gonur-1, where the exca¬
vator reported a variety of “altars,” two of which showed signs of the hallucinogenic
ephedra when subjected to microscopic analysis (Sarianidi 1990, 1993b, c). Poppy pol¬
len was also identified on pestles and grinding stones of the site.

Opposite these “altars” were “fire altars,” or, perhaps more accurately, rooms that
shewed the effects of fire. Summing up the archaeological evidence, Sarianidi (1930c)
concludes, “It can be said with a high degree of certainty that Togolok-21 must have

Archaeological Evidence outside the Subcontinent 213

The BMAC (Bactria Margiana Archaeological Culture).

been a temple connected in some or other way with a religious cult during which ven¬
erations of the sacred fire and libations took place” (252). Chemical analysis performed
upon vessels discovered in a similar site at Gonur determined that the vessels had con¬
tained cannabis, ephedra, and poppy. Also of relevance at this site was the discovery of
ceramic stands and sieves “by means of which the juice was divided from the solid pressed-
out mass” (252). If one is to follow, for example, Nyberg (1995), then ephedra is the
most likely candidate for the soma/hoama cult of the Indo-Aryans and the Iranians. This
is certainly the connection that Sarianidi makes. In short, he holds that “the Vedic Aryans
were part of the BMAC tribes whose culture replaced the Harappan” (1993c, 263). Nyberg
(1995), however, subjected the samples from these sites to pollen analysis at the Uni¬
versity of Helsinki but could not find any trace of ephedra (although Sarianidi, 1999,
has questioned this analysis).

While all this is well and good, attention must be drawn to the fact that the BMAC
was a sophisticated civilization consisting of fortified towns. The temple structures just
noted were “monumental”; the Gonur temple occupies an area of two hectares (from a
total area of twenty-two hectares) and was surrounded by walls up to four meters thick
(Sarianidi 1993b, 8). Indeed, according to later excavators of BMAC sites:

The extensive distribution of the BMAC, together with the recognition of the consider¬
able size of their cities . . . and die monumental nature of single architectural units, . . .
combine to suggest that we are dealing here with a socio-political phenomenon of consid¬
erable magnitude . . . that involved a substantial region of Middle Asia. ... It is our
belief that the BMAC, if not itself a state, appears to mirror a facsimile of state-structured
polity of power. The 15 km. distance from Gonur to Togolok in Margiana is no greater
than the distance which separated Lagash from Girsu in Mesopotamia. Nor is there a
great difference in the relative size of these respective cities. The latter are taken, without

214 The Quest for the Origin of Vedic Culture

qualification, to be important “city-states” within Mesopotamia. (Hiebert and Lamberg-

Karlovsky 1992, 11)

As discussed in the last chapter, scholars since the time of Sir John Marshall have dis¬
allowed the possibility of an Indo-Aryan presence in the Indus Valley by insisting that
the social horizon of the Rgveda knows no urban settings and is purely pastoral. Rau’s
depiction of the Vedic pur is one of wattle huts. 1 will question later whether this posi¬
tion is tenable based on the fact that settlements and towns continued to exist in the
post-Harappan period, albeit shifting from the Harappan nucleus, at a time when no
one can doubt that the Indo-Aryans were very much situated in the region. Lack of
urban references in the Rgveda may not be a reliable indicator of the Indo-Aryan’s igno¬
rance of urban centers.

The same logic is applicable here. If the DMAC culture was Indo-Aryan, then the
Indo-Aryans certainly knew and lived in (and, according to some scholars, established)
urban centers in a state-like context. Anyone accepting the Indo-Aryan identity of the
BMAC (or any of die sites on the southern routes) cannot dien deny the possibility of the
Indus Valley Civilization possibly being an Indo-Aryan civilization on diese particular
grounds, namely, the apparent lack of urban references in the Rgveda. Even if it is argued
that the Indo-Aryans were not the founders of this civilization, but arrived toward the end
of the Bactria and Margiana cultural complex (drey were initially held to have destroyed it
until it was realized that no traces of destruction have been found), they nonetheless
must surely have passed dirough this whole area on their way to India and so must
have been aware of, and interacted with, such towns. Accordingly, it becomes very hard
to deny the possible residence of the Indo-Aryans in, or coexistence with, urban centers.

Also of relevance to the last chapter (where we found that the principal reason given
for the non-Indo-Aryan identity of the Indus Valley is the absence of horse bones there)
is the fact that although the horse was certainly utilized in the BMAC, as attested by its
representation on grave goods, no horse bones have been discovered there (despite the
fact that an unusually high number of animal bones have been found). So we have
proof of horse, but no horse bones. Absence of horse bones, then, may not equal ab¬
sence of horses, nor necessarily of Indo-Aryans. A final related point is that the Vedas
make no mention of temples, or temple structures. What are we to make of “ceremo¬
nial centers” of the sites of Togolok-21 and Gonur-1 discovered in the same BMAC
where we have fire worship and the ritual usage of hallucinogens in a templelike set¬
ting? Any acceptance that all this was the handiwork of Indo-Aryans will entail aban¬
doning certain stereotypes such as that the Indo-Aryans knew no urban centers or temples
and that the failure of archaeologists to uncover horse bones equals the real-life absence
of horses in a society.

Hiebert and Lamberg-Karlovsky disagree with Sarianidi concerning the external ori¬
gin (from southeastern Iran) of the BMAC, preferring to consider it “the development
of a new type of social structure within an ongoing culture, rather than migration or
invasion” (Hiebert 1995, 200). This position has been accepted by Tosi (1988), albeit
in conjunction with a “massive immigration of external elements” (62). Hiebert and
Lamberg-Karlovsky note that there is no site in southeastern Iran providing any evi¬
dence for such a claim, and that any BMAC material found on sites from that area is
intrusive in nature. In other words, if there is a movement, it is from east to west. These

Archaeological Evidence autude the Subcontinent 2 1 5

scholars are prepared to consider, however, that the BMAC culture is Indo-Aryan. Of
particular relevance is the intrusion of burial assemblages with artifacts typical of die BMAC
culture into the Iranian plateau and the western borders of the Indus Valley in the sites
of Mehrgarh VII and Sibri in Baluchistan, which “may be correlated with the introduc¬
tion of the Indo-European language” (Hiebert and Lamberg-Karlovsky 1992, l).

While acknowledging that “we must distinguish between the movement of peoples
from the movement of objects and/or styles by exchange and/or stimulus diffusion,”
these scholars are of the opinion that the “evidence suggests that the people buried in
the tombs, with an exclusively Central Asian [BMAC] material inventory, came from
Central Asia” (Hiebert and Lamberg-Karlovsky 1992, 3). However, they also note that
the tombs with such exclusive material are “rare." Nonetheless, here we do seem to
have the first evidence of an archaeological intrusion into the subcontinent from Cen¬
tral Asia during the commonly accepted time frame for the arrival of the Indo-Aryans:
“Since the BMAC is clearly intrusive on the Iranian Plateau and in the hill country of
Baluchistan, we suggest that the BMAC provides the first archaeological evidence that
meets both the chronological and historical requirements for the introduction of the
Indo-Iranian language onto the Iranian Plateau” (10). Kohl (1984) likewise accepts
“the possibility of a movement of peoples to the south, possibly displaced by steppe
and mounted nomads from further north” (242). Other scholars have also correlated
the apparent intrusiveness of these graves with the lndo-Aryans (Parpola 1994, 147;'
Allchin 1995, 47). Scholars also point to the similar pottery sherds found in nearby
Nausharo, connecting this site with Mehrgarh VIII and Sibri, as further evidence of
such incursions.

The Mehrgarh and Sibri cemeteries, excavated by Jean-Francois Jarrige, yielded “abun¬
dant material presenting obvious parallels with East Iran (Shahdad), Northwest Afghani¬
stan (Dashly) southern Uzbekistan (Sappali) and the Murghab region” (Jarrige and
Hassan, 1985, 150). Ceramic objects and an amulet also “indicate some degree of
contemporaneity and interaction between the Mehrgarh VH/Sibri complex and the Indus
civilization” (150). There is debate regarding the origin of these finds. Lamberg-Karlovsky
(1993) points to two possibilities—“The movement of peoples from Central Asia . . . or,
in the absence of such a migration, the presence of strong culmral influences uniting
these two areas” (35)—but favors the former hypothesis. However, Jarrige (1989), the
excavator of the site, is less inclined to see these finds as evidence of population move¬
ments: “The evidence of a formative period of the cultural complex of Mehrgarh VIII/
Sibri at Nausharo . . . cannot be interpreted in term of invasions from the north-west to
the south-east but within the framework of fruitful intercourse at a time when Mohenjo-
daro is still an active city” (67). Moreover, there is no evidence of the horse at Sibri
(Santoni 1985), which scholars have long insisted must accompany any proposed Indo-
Aryan identification.

Along the same lines, the antiquities discovered at Quetta in 1985, which are also
sometimes connected with intruding Indo-Aryans (i.e., e.g. Allchin 1995), can also sim¬
ply be viewed as reflecting ’’the economic dynamism of the area extending from South
Central Asia to the Indus Valley.” The fact that similar objects are also found in graves
and deposits in northern Iran, eastern Iran, northwestern Afghanistan, South Turkmenia,
and Baluchistan might simply indicate “a wide distribution of common beliefs and ritual
practices” ( Jarrige and Hassan (1985) 1989, 162-163). Jarrige and Hassan reject the

216 The Quest for the Origin of Vedic Culture

idea that these finds were associated with invaders related to the Hissar Ill C complex,
since “there is nothing in the Gorgan Plain and at Hissar to prove that northern Iran
has been a relay station for invading people. The . . . grey ware can very well be ex¬
plained within its local context” (163-164). Nor are these scholars partial to the north¬
ern steppe Andronov alternatives, since:

We leave to the linguists the problem of whether Indo-European languages were introduced
into the Middle Asian regions from a still unknown part of the Eurasian steppes in the
course of the third millennium or if Indo-lranian languages have been associated with
these regions for a much longer period. As far as archaeology is concerned, we do think
that it is increasingly necessary for specialists in Indo-lranian studies to pay attention to
the . . . interrelated cultural entities of the late third and early second millennium in the
regions between Mesopotamia and the Indus. It is a direction of research that is likely to
be more fruitful than are traditional attempts to locate remains left by nomads from “the
Steppes,” attempts that were in fashion when the Indo-lranian Borderlands were thought
to be a cultural vacuum. (164; my italics)

Despite inviting linguists to reconsider the northern steppe hypothesis in favor of the
southern route, it can be inferred from Jarrige and Hassan, as from the work of a num¬
ber of archaeologists considering the problem of Indo-Aryan origins, that the Indo-Aryan-
locating project exists solely due to linguistic exigencies:

The development of original but closely interrelated cultural units at the end of the third
and the beginning of the second millennium cannot be explained just by the wandering
of a single group of invaders. The processes were obviously multidirectional in regions
with strong and ancient cultural traditions. This does not preclude the fact that move¬
ment of population and military expeditions . . . may have played an important historical
part but, as far as archaeology is concerned, there is nothing to substantiate a simplistic
model of invasion to account for the complex economic and cultural phenomena mani¬
fest at the end of the third millennium in the regions between Mesopotamia and the
Indus Valley. (164)

Nonetheless, those who do find the aforementioned linguistic exigencies compelling
must find some way of getting the Indo-Aryans speakers into the subcontinent by some
means or another. Mallory (1998) feels comfortable enough ascribing some form of Indo-
lranian identity to the Andronovo culture but admits that, “on the other hand, we find
it extraordinarily difficult to make a case for expansions from this northern region to
northern India . . . where we would presume Indo-Aryans had settled by the mid-sec¬
ond millennium bce” (191). Referring to the attempts at connecting the Indo-Aryans to
such sites as the Bishkent and Vakhsh cultures, he remarks that “this type of explana¬
tion only gets the Indo-lranian to Central Asia, but not as far as the seats of the Medes,
Persians or Indo-Aryans” (192). He points out that suggesting an Indo-Aryan identity
for the BMAC requires a presumption that this culture was dominated by steppe tribes.
However, “while there is no doubt that there was a steppe presence on BMAC sites,
. . . this is very far from demonstrating the adoption of an Indo-lranian language by the
Central Asia urban population” (192),

Mallory (1998) offers a Kulturkugel (culture bullet) as a possible explanatory model
for the Indo-Aryan incursions, although remarking that “German is employed here
to enhance the respectability of an already shaky model” (192). This conceptual pro-

Archaeological Evidence outside the Subcontinent 217

jectile is envisioned as an Indo-Iranian linguistic bullet propelled by the social orga¬
nization of the steppes outlined previously and tipped with a nose of malleable
Andronovo material culture. After impacting the BMAC culture, the projectile con¬
tinues on its trajectory, but now as an Indo-Aryan linguistic bullet with a BMAC cultural
tip. In other words, the steppe tribes entered the BMAC, shed the trappings of their
Andronovo heritage, and then, reacculturated, continued on their way toward India
after having adopted the cultural baggage of the BMAC and undergone the linguistic
transformations separating the language of the Indo-Iranians from that of the Indo-
Aryans. Mallory is too good of a scholar nor to immediately include an addendum,
stating that “the introduction of the kulturkugel emphasizes the tendentious nature of
any arguements for the dispersals of the Indo-Iranians into their historic seats south
of Central Asia” (193). He is also candid enough to point out that “it is . . . difficult
to imagine how such a concept could be verified in the archaeological record or, to
continue the metaphor, could be traced back to the original ‘smoking gun’” (194).
Mallory’s Kulturkugel is the type of gymnastics incumbent on anyone atempting to
find archaeological evidence of the Indo-Aryans all the way across Asia and into the
subcontinent.

Two Wave Theories

The idea of a two-wave (or multiple-wave) incursion into India has been suggested sev¬
eral times by linguists for over a century and has entered the realm of archaeology more
recently through the work of Asko Parpola. Hoernle (1880) laid the genesis of the idea
in 1880 based primarily on the phonemic and morphological features that distinguished
the New Indo-Aryan languages into two main groups, which he termed the “Sauraseni
Prakrit tongue” (Sr. Pr.) and the “Magadhi Prakrit tongue” (Mg. Pr.). From a historical
perspective, he envisioned this situation coming about as follows: “The Mg. Pr. and the
Pashtu and Kafiri were once in close connection, perhaps one language. ... at some
time in the remote past, they became separated by the Sr. Pr. tongue, like a wedge, cleaving
diem asunder and gradually pushing the Mg. farther and farther away towards the east”
(Hoernle 1880, xxxiv).

Grierson (1903) took Hoernle’s wedge concept and expanded it further. He envi¬
sioned the Aryan invasion taking place over centuries in multiple waves. Using the first
and the last of such waves as reference points, he proposed that the earlier comers spoke
a non-Sanskritic Indo-Aryan dialect, 16 and the newcomers a Sanskritic one: “The later
invaders . . . reached the Punjab which they found already settled by Indo-Aryans from
the West speaking a closely cognate tongue” (52) From there, they forced their way to
the eastern Punjab, which they wrested from the first comers, pushing them to the sur¬
rounding areas: “It is reasonable to suppose that the tribes who composed this later
invasion (wherever they came from) should have expanded as time went on, and should
have dirust outwards in each direction the members of the earlier incomers” (53) Grierson
observes that in medieval Sanskrit geography, Madhyadesa is continually referred to as
the true, pure home of the Indo-Aryan people. He further notes that the modern San¬
skritic Indo-Aryan vernaculars fall into two main families, one of which is spoken in a
compact tract of country almost exactly corresponding to this ancient Madhyadesa and

218 The Quest for the Origin of Vedic Culture

centered in modern-day Uttar Pradesh. The other surrounds it in three-quarters of
a circle, commencing in Kashmir and running through the western Punjab, Sind,
Maharashtra, central India, Orissa, Bihar, Bengal, and Assam (52-53). These two ar¬
eas represent the offspring of the two waves of Aryan incursions.

Asko Parpola (1988, 1993, 1994, 1995, 1997) has offered the most elaborate attempt
to identify a two-wave Indo-Aryan incursion in the archaeological record. Parpola at¬
tempts to harmonize the discrepancies of associating the Indo-Aryans with the Andronovo
culture as well as with the BMAC one by proposing a twofold incursion of Indo-Aryans
into the subcontinent. The first of these—the Dasas of the Rgveda-were Andronovo-
related tribes that took over the BMAC. The sudden upsurge of wealth of the BMAC
around the twenty-first century is likely to have been due to the strong hierarchical in¬
fluence of this ruling elite. Moreover, he points out that, the BMAC occupies precisely
the area around the Oxus where Persian and Greek historical sources place the Da(h)as
(< Dasas). The Parna (< Panis), another Rgvedic tribe inimical to the Aryas, are also
identified in these classical sources as coming from this area.

This first wave of Indo-Aryans was engulfed by later soma-pressing Indo-Aryan
Andronov tribes that eventually became the composers of the Rgveda wherein they re¬
fer to themselves as Aryas. Their arrival in this area may well have been the cause for
the sudden collapse of the BMAC around 1700 b.c.e. This date synchronizes neatly
with the Gandhara Grave Ware culture (which will be discussed in the next chapter),
the first unambiguous appearance of the horse in the subcontinent, as well as with the
subsequent appearance of the Mitanni in Syria around 1500 b.c.e. By this time, the
Aryas and Dasas had merged as evidenced by the juxtaposition of the Aryan Indra,
with Varuna of the Dasas in the Mitanni treatises. Parpola’s method, to some extent,
involves determining the route the Aryans must have taken in order to reach India,
examining the archaeological record along this route, and composing an elaborate his¬
torical scenario that he believes accords with Rgvedic narrative: : “The valley of Swat
occupies a strategic position in the archaeological identification of the early Rgvedic Aryans,
because they must have passed through this area. . . . Therefore we must briefly review
its archaeological history to check the match" (Parpola 1994, 153; my italics).

Parpola (1988) extracts the Rgvedic verses where Indra, the purandara ‘fort-destroyer’
is active in the destruction of the ninety-nine or hundred puras ‘forts’ of his enemies,
the Dasas. 17 Based on a verse from the Satapatha Brahmana (6.3.3.24-25), and draw¬
ing on the work of Wilhelm Rau discussed earlier, Parpola proposes that a significant
feature of these forts is that they are tripura, or have a threefold structure. Parpola takes
this to mean not only that the forts are surrounded by concentric circular walls but also
that the forts themselves are circular in construction. Although only one such round
fort from this period has been unearthed to date, Parpola believes that there must have
been more, since circular forts with concentric walls survived there until Achaemenid
times. He argues that both of these details preclude the possibility of the cities of the
Indus Valley representing the Dasa forts, since these are neither round nor fortified by
three concentric walls. TheRgveda, however, is firmly situated in more or less the same
geographic area occupied by the Indus Valley, so if the Dasa forts were not those of the
Harappans, whose were they?

Parpola argues that even though the famous Sudas of the Rgveda is clearly associated
with the Punjab, 18 his forefather, Divodasa, need not be. Divodasa’s chief enemy.

Archaeological Evidence outside the Subcontinent 219

Sambara, possessed a hundred (or ninety-nine) forts and was said to have resided in a
mountainous domain. This, according to Parpola, could have been Bactria, northern
Afghanistan. As has been discussed, there is a site in this area called Dashly-3, which is
dated to about 2000 b.c.e. The site, although surrounded by square walls, consists of
various buildings, among which stand three concentric walls. Although these urban
structures are a far cry from the temporary mud and wattle purs reconstructed by Rau,
these three walls correspond, for Parpola, to the tripura of die Dasas. He believes he has
found the evidence representing the Dasa forts attacked by the Aryans on their way to
the subcontinent. 19

K. D. Sethna, who is questioning the very assumption that the Indo-Aryans need be
considered intruders at all, has dedicated half of the second edition of his book The
Problem of Aryan Origins (1992) to critiquing Parpola’s 1988 . article, which had been
published in the interim between editions. Sethna’s general objections concerning in¬
terpretative methodology echo those of many South Asian archaeologists: “The picture
we derive from Parpola is of a traffic to and fro of cultural modes—continued from a
fairly long past and across sufficiently wide areas—against a common religious background
of various shades. It is a picture of contacts and exchanges. . . . none of them necessar¬
ily bespeak large-scale movements of population” (229). Since all the specific objections
against each step of Parpola’s reconstruction are quite voluminous and painstakingly
argued, one example (central to Parpola’s theme) of how Sethna sees the “evidence”
being artificially construed to fit a series of assumptions will suffice for our purposes.

Sethna finds the passage from the Satapatha Brahmana upon which Parpola bases
his case of concentric walls actually describing the gods as fearful lest the Raksasas, the
fiends, might slay their Agni. 20 The passage explains why the priest draws three concen¬
tric lines around the fire during a particular rinial. The practice, mentioned in two other
places in the Brahmanas, is enacted to ward off demons during the performance of the
sacred rites. Sethna points out that, first of all, the three walls represent fire ( agnipurd ),
not stone and mortar. Parpola has reified a magico-ritualistic ceremony into a real-life
fortification. Second, it is the Aryan sacrificers who are drawing the sacred lines (or
building the forts as per Parpola), not the Dasas or Asuras, as Parpola’s version requires.
Moreover, neither the word tripura nor any of its associations mentioned earlier occur
anywhere in the Rgveda itself. As we have seen, in the Rgveda, the puras are described
variously as wide and broad, made of stone, made of metal, hundred-curved, or with
the strange epithet ‘“autumnal,” but there is no mention whatsoever of three concentric
walls. 21 One might add, too, that in this account Sudasa and his father Divodasa, as is
obvious by their names, are themselves Dasas, despite being unquestionably Aryans
par excellence, which throws the whole Arya-Dasa dichotomy into even further confu¬
sion. Parpola sees the Dasas as a first group of invading Indo-Aryans, displaced by a
second group who called themselves Aryas and their foes, the first group, Dasas. This
raises the question of why Sudas and his father, who were leaders of the second, victo¬
rious group, would call themselves, or have been called, Dasas (Parpola accounts for
this by supposing that the Vedic group had adopted Dasa traits from early on).

Sethna further points out that the three concentric walls of Dashly-3, even if we al¬
low them to be the real-life protoforms of the protective fire lines drawn by the priest in
the much later Vedic rituals, are a single archaeological occurrence. No other fortified
village in Bactria has three concentric walls (we can recall that Sambara alone, whom

220 The Quest for the Origin of Vet! ic Culture

Parpola considers to be one of the leaders of the Dasas in their concentric forts in Bactria,
is described as having ninety-nine forts). Nor, as Parpola (1988, 138) himself notes, is
there any evidence that any of these fortified villages were burned, unlike those described
in the Rgveda. Moreover, the solitary site that does have three concentric walls does not
stand out on its own, as Parpola’s conjectured Dasa forts should, but occurs inside the
square walls surrounding the fort. It would have been these square walls that would
have been visible to any invading Aryans, not the concentric ones of Parpola’s account.
Sethna (1992) continues at length in his criticisms and concludes: “I do not think we
have any reason to visualize or locate Dasa forts in the way Parpola does” (313).

Parpola’s account has received criticisms from various other quarters. Sarianidi (1993b)
notes:

It should be indicated that the available direct archaeological data contradict the theory,
suggested long ago, concerning the intensive penetration of the steppe Andronovo-type
tribes into traditional agricultural areas. Direct archaeological data from Bactria and
Margiana show without any shade of doubt that Andronovo tribes penetrated to a mini¬
mum extent into Bactria and Margianian oases, not exceeding the limits of normal con¬
tacts so natural for tribes with different economical structures, living in the borderlands
of steppe and agricultural oases.” (1 7)

Lyonnet (1993) also points out that “no traces of systematic destruction, at least in
Bactriana and Margiana, have been observed and nothing else allows us to state that
there was an intrusion of invaders. This intrusion of new objects could be considered
the result of peaceful trade, and the new rites could be due just to local change” (428).
Like Sethna, she also draws attention to the fact that Dashly-3 is die only circular fort
with three surrounding walls that has been found, and to the fact that the proto-urban
aspect of the site does not accord well with the generally accepted notion of the Rgvedic
descriptions which are suppossed to know no urban centers. Moreover, Jettmar (1981)
observes that the “circular walls had no value of defence,” since houses lean on both
sides of them directly and they “were rather thin” (222). Parpola himself is prepared to
“admit that concrete evidence of an invasion of steppe nomads into Bactria and Margiana
is missing” (1993, 55). While not sharing the same views as Parpola, I would like to
agree with Sarianidi (1993b) that “you can only admire A. Parpola having taken on
gigantic effort and his having made an attempt to logically unify archaeological and lin¬
guistic data. ... no archaeologist has dared to undertake such a thankless and titanic
work” (18).

Conclusion

From the perspective of archaeology, then, we have two basic options. We can either
accept (or renegotiate) the temporal and geographic contours assigned to the trajectory
of the Indo-Aryans from a non-South Asian origin and attempt to correlate their move¬
ments into the subcontinent with some of the innovations in the archaeological record,
both in central Asia and into the subcontinent. Or we can challenge or (as with many
Indigenous Aryanists) simply ignore all these assumptions and interpret the continu¬
ities and innovations in the archaeological record as nothing more than regular exchanges
and interactions among various ethnic and socio-economic groups without feeling im-

Archaeological Evidence outside the Subcontinent 221

pelled to connect them with the physical movements of a new language group (this is
not to suggest that most Migrationists do not also accept such exchanges and interac¬
tions). Of those following this latter course, some are simply unconcerned with the lin¬
guistic evidence and operate purely as archaeologists, while others hold that the Indo-
Aryan languages have always been indigenous to the subcontinent. All these positions
have corollaries.

If we adopt the first Migrationist position, we need to orient ourselves around either
a northern course or a southern course (or some combination of both). The most seri¬
ous, obvious, and oft-cited objection against the northern Andronovo course is that the
steppe culture does not intrude into the South Asian borderlands (not to speak of the
heartland). Why, then, should one accept it as representing Indo-Aryan speakers in¬
truding into South Asia (although these steppe people may certainly have been speakers
of Indo-Iranian dialects)? Such a position can only be predicated on an acceptance of
the linguistic assumptions outlined in the previous chapters and not on the archaeo¬
logical data per se.

In response to this, one can try one’s hand at mixing and matching and propose
that the Indo-Aryans adopted significantly different local cultures as they moved along
or suggest that they arrived in separate waves. Or, since “in the northwest zone there is
a striking absence of uniformity and common traits in any of the cultures identified as
‘Aryan’” (Dikshit 1985, 57), one can choose to find traces of Indo-Aryans in every place
where any innovations are to be found at all. Allchin finds Aryans at Hissar III, at
Namazga, in the cemeteries of Tajikistan, in the BMAC, in Mehrgarh, Sibri, and Quetta,
as destroyers of the Mature Harappan cities, as residents of the Mature Harappan cities,
in Cemetery H, in Chanhu-daro, and in the Gandhara grave culture. This entails sub¬
scribing to a “model flexible enough to allow for several different types of movement,
probably taking place on more than one occasion, and representing a number of stages
in time and place” (Allchin 1995, 47). Mallory’s “bullet” is a somewhat less ambitious
version of this method. Such arguments, however, cannot be premised on the archaeo¬
logical data. They can only be supported by philological predispositions imposed upon
whatever archaeological evidence appears amenable or most readily at hand. As far as
B. K. Thapar (1970) is concerned:

The archaeological and the anthropological evidences, represented by the various culture-
groups of the second millennium B.C., are inconsistent with the philological evidence. Even
the archaeological and anthropological evidences have been found to vary from region to
region—Anatolia, northern Iraq, northern Iran, Soviet Central Russia, Swat valley and
Gandhara region or Pakistan and Ganga-Yamuna doab. ... It is obvious, therefore, that
there was no single culture associated with the Aryans in all these regions. . . . Are wc to
assume that the Aryans were migrants with no defined culture but with adherence to a
linguistic equipment? (160)

Dilip Chakrabarti’s comments (1986) are of relevance here: “Archaeology must take
the entire basic framework of the Aryan model into consideration. It should not be a
question of underlining a particular set of archaeological data and arguing that these
data conform to a particular section of the Vedic literary corpus without at all trying to
determine how this hypothesis will affect the other sets of the contemporary archaeo¬
logical data and the other sections of the Vedic literary corpus” (74). As far as Thapar
and Rahman (1996) are concerned, there is nothing “to show that the Indo-Aryan peoples

222 The Quest for the Origin of Vedic Culture

possessed a specific material culture, special pottery, or particular figurines that would
enable us to establish some kind of identification marks for their migration” (1996,
277). Edmund Leach (1990) rejects tire whole method of using the Rgvedic literary corpus
to try to extract clues that can be correlated with the archaeological record. He objects to
the entire procedure whereby “an oral tradition has been treated as if it were a datable
written record and myth has been confused with history as if it actually happened” (230). 22
Leach throws out an appeal to his colleagues: “If only . . . [scholars] would stop think¬
ing of the Rig Veda as a garbled history book” (244). As far as he is concerned, “The
Aryan invasions never happened at all”, but, he adds, “of course no one is going to
believe that” (245).

The immediate objection against the southern course is that the Indo-Aryans are
primarily considered to have been pastoralists who supposedly knew no urban centers
(and what to speak of states) or monumental places of worship, and were not agricultur¬
ists. I have already indicated my position that the Indo-Aryans must have interacted
with urban centers whenever and wherever they were. Moreover, it has been pointed
out that agriculture is very old on the subcontinent, and so no pastoral group could
have existed there oblivious to farming cultures. Therefore, such objections need quali¬
fications.

From Indigenist perspetives Pastoralists and farmers have always coexisted in the
subcontinent. Both pastoralists and agriculturalists presently are, and have long been,
Indo-Aryan speakers. But how far can we go back in India with an Indo-Aryan pres¬
ence? Have the Indo-Aryans, whether agriculturists, pastoralists, or both, been in the
subcontinent since all eternity? The immediate issue confronting an Indigenous posi¬
tion is that if the Indo-Aryan speakers are autochthonous to South Asia, then how are
they related to the other Indo-European languages? The corollary can only be that the
other Indo-Europeans left the subcontinent for their destinations to the west.

This raises the immediate objection that if archaeology cannot trace any consistent
material culture identifiable as Indo-Aryan arriving into the subcontinent from outside,
it most certainly cannot identify any such culture emanating out. Accordingly, as far as
archaeology is concerned, we have reached a stalemate (although from a Migrationist
perspective there is, arguably at least, some kind of chronological sequence of archaeo¬
logical culture that at least heads toward the general direction of the subcontinent, even
if it does not penetrate it). Ultimately, however, the Aryans cannot be satisfactorily iden¬
tified in the archaeological record as either entering or exiting. The trajectory of the
Indo-Aryans, indeed the necessity of their very existence, is a linguistic issue that ar¬
chaeology, as most archaeologists are well aware, cannot locate in the archaeological
record without engaging in what, to all intents and purposes, amounts to special and
often complicated pleading. On the basis of the present evidence, linguistics cannot
decisively determine with any significant degree of consensus where the original home¬
land actually was. And archaeology can only hope to be productive in identifying the
material remains of a linguistic group if linguistics has already done the groundwork of
pinpointing its geographic area of origin with a reasonable degree of precision.

Accordingly, archaeology cannot deny the possibility that Indo-Aryan and Iranian
(which were preceded by Indo-Iranian) languages might have been spoken in the area
of the Punjab, Pakistan/Afghanistan, southeast central Asia/northeast Iran since the
second, third or even fourth millennium b.c.e. The problem is chronological. In fact,

Archaeological Evidence outside the Subcontinent 223

archaeologically at least, South Asian archaeologists often draw attention to a cultural
continuum that can be traced as far back as Mehrgarh in the seventh millennium b.c.e.
within which innovations and developments can be explained simply by internal devel¬
opments and external trade. If there were no constraints stemming from the date com¬
monly assigned to the Veda, this whole area could have included urbanites and agricul¬
turists from the South, as well as nomads and pastoralists from the North, interacting
together in the millennia b.c.e. as they always have been and still do in the present day.
Both steppe dwellers and urban farmers could have been speakers of related Indo-Iranian
dialects in protohistory just as they are today and have always been in recorded history.
There could have been invasions, migrations, trade, cultural exchanges, all manner of
interactions—cultural evolution and devolution (followed sometimes by renewed evolu¬
tion)—as well as all manner of diversification in chronological time. And all within a
large, heterogeneous ethnic and cultural area of people who nonetheless spoke related
dialects—whether living in towns, mountains, or agricultural plains—just as has always
been die history of the subcontinent. Sarianidi (1999), for one, is prepared to countenace
something along these lines in the second millennium bce (albeit from a Near Eastern
homeland):

The northern branch of the Indo-Iranian tribes had separated early on from its Indo-
Iranian mother country and settled widely in the wood and steppe zone of Eurasia . . . We
find this reformed culture in the Andronovo tribes in the middle of the second millen¬
nium bc in Central Asia where they contacted the Indo-Iranian population of the farm¬
ing oasis of the BMAC . . . And maybe for this reason the Andronovo tribes found no
confrontation in the territory of the BMAC (not to speak of a military invasion). On
the contrary they peacefully co-existed with the ancient farmers of Margiana, south
Turkmenistan and Bactria as they were all tribes of related origin. (323)

It is not archaeology that can discount the possibility of all this. As I have asserted
elsewhere, everything depends on die date of the Rgveda. Before turning to the prob¬
lems involved with the dating of this text, there is one last aspect of the archaeological
record that must briefly detain us. This is the attempt to trace the trajectory of the Indo-
Aryans from within the subcontinent.

Aryans in the Archaeological Record

The Evidence inside the Subcontinent

Once it had been determined and accepted by a majority of scholars that the Indus
Valley Civilization was definitely not to be connected with the Vedic Aryans, and that
the Vedas were to be dated sometime around 1200 b.c.e., the time span for the main
entry of the Indo-Aryans into the subcontinent could be narrowed down to a post-
Harappan period between approximately 1900 and 1200 b.c.e. I have discussed how
the archaeological cultures from the Caspian Sea area to Afghanistan in the particular
time frame deemed appropriate for the trans-Asiatic saga of the Indo-Aryans were scoured
and how archaeologists interpreted a variety of material cultures as evidence of Indo-
Aryans, either by attempting to correlate them with written sources or on the basis of
their innovatory nature in the archaeological record. A parallel logic has prompted a
series of interpretations from within the Indian subcontinent and within this later time
frame.

The immediate and most obvious archaeological candidates were the archaeological
cultures known variously as Late Harappan, Post-Harappan, or Post-Urban, during which
period the Indo-Aryans are assumed to have entered the subcontinent. Allchin (1995,
29), opting for Possehl’s term of Post-Urban, dates this phase from 2000 b.c.e. to the
last quarter of the second millennium b.c.e. Allchin notes a variety of factors that con¬
tributed to the weakening of the urban Indus culture, which he suggests “presented an
invitation to some of their predatory neighbours, and thus coincided with incursions of
peoples from the hills to the west, who almost certainly established for themselves
dominant roles in relation to the existing population of the Indus system” (38). A num¬
ber of regional cultures flourished during this period, from which I will consider the
ones that have been typically connected with intruding Indo-Aryan elements. These are
the Gandhara Grave Culture; the Jhukar Cultures, the Cemetery H culture, and the

224

Archaeological Evidence inside the Subcontinent 225

Painted Gray Ware culture. I will first discuss the ways in which these cultures have
been related to the Indo-Aryans and then discuss interpretative options. Again, 1 will
attempt to spare the reader excessive detail and simply touch upon the arguments likely
to be most commonly encountered in this regard.

Gandhara Grave Culture

A variety of graves were found in the environs of the Swat Valley by A. Dani, who
coined the name for this culture, and by members of the Italian Archaeological Mis¬
sion. 1 Some of the graves are of fractional skeletons, 2 which, according to Dani (1992),
are the only graves associated with iron “and must be understood as an intrusive phe¬
nomenon by a people who introduced iron” (405). In one graveyard, a burial of two
horses was found at surface level (Antonini 1973, 241), which, as we know, will imme¬
diately capture the attention of Aryan-spotters. In addition to a red ware pottery, there
is a plain grayware that belongs to a tradition “very different from those of the periods
immediately preceding and immediately following, in shapes and decoration and in the
production techniques of the vessels” (Stacul 1973, 197). Dani (1978) and Stacul (1969)
connect this gray pottery with die plain ware, from northeast Iran, noted previously. 3
Various bowls, vases, and bronzes similar to objects from Hissar III fortify this associa¬
tion. Dani finds close relations between the material culture of the Northwest and that
of the northeastern part of Iran in the second and first millennium B.C., but with little
direct link with the areas in Soviet Central Asia or western Asia. According to him, this
part of Iran, lying to the east of the Caspian Sea, should be regarded as a nucleus zone
for the diffusion of cultures to different directions. The culture he sees diffused from
this area is that of the Indo-Aryans: “The literary accounts have talked of the people
who call themselves Aryans in the Vedic literature. ... as far as we are concerned, we
. . . have shown how in the three periods of the graves we should now understand the
literary records which we have so long taken for historical truth” (Dani 1978, 53).

Stacul (1969) finds the excavations can be assigned to seven quite distinct periods,
although “we are unable to define just how much such clear differences between period
and period depend on the fragmentary nature of our data . . . and how much is the
result of sudden changes and upheavals due to the instability of the dominant tribal
groups” (86). Unlike Dani, he does connect such groups with the central Asian peoples
he believes were involved in the decline of the urban centers in Afghanistan, northern
Iran, and central Asia. But he notes that “the complexity of these phases and the gen¬
erally scarce archaeological documentation hinder even a summary reconstruction of
these events, which should presumably be linked to the spread of the Aryan tribes in
the regions of the Indo-Pakistan sub-continent” (86-87). 4

For those undertaking the mix-and-match project that is the inevitable lot of anyone
attempting to identify the Indo-Aryans in the archaeological record, there is no consen¬
sus among archaeologists regarding any possible foreign affiliation of the graves. Thapar
and Rahman (1996) find “their connection to central Asia and Iran are definite, though
of what character is still not known” (268). Kuzmina (1976) is much more confident
that the burial rites and pottery from both the Beshkent and the Gandharan cemetery
provide “grounds for attributing an Indian origin [to the Beshkent group] and for sup-

226 The Quest for the Origins of Vedic Culture

posing that it later shifted to Hindustan where the Swat cemeteries are thought to have
been its legacy” (131).

Antonini, in contrast, has rejected attempts to correlate the graves with the Beshkent
and Vakhsh graves of Tadjikistan (for references to these attempts by scholars such as
Kuzmina and Litvinskij, see Antonini 1973). Antonini (1973, 239-244) points out
differences in tomb structures, metal objects, and pottery types between the two cub
tures, as well as the absence in the Swat graves of items that were typical of the
Tadjikistan graves, such as male-female differentiation in inhumation, miniature
hearths, stones arranged as swastikas, and animal remains. Although both Dani and
Stacul accept some Iranian parallels with the grave material, Dani (1977) finds “very
little” that provides a link with central Asia, while Stacul also notes that the Tadjikistan
graves “do not seem to indicate affinities with the graves found in Northern Paki¬
stan” (198). In terms of other connections, Parpola (1993), following Mallory, accepts
that this culture is “by no means identical with the Bronze Age Culture of Bactria
and Margiana” either (54).

Tusa (1977) believes that “the so-called ‘grave culture’ is not in fact due to a sudden
interruption in the life of the valley but to an appreciable, substantial change perhaps
due to new contributions that are nevertheless in line with the cultural traditions of the
previous period.” He echoes objections that have been raised so many times by South
Asian archaeologists: “The existence of contributions from the outside, for too long used
to justify cultural change in the sub-Himalayan area, has in my opinion been exagger¬
ated even though it could conceivably have been a factor in cultural change without
being the only one” (690). As far as he is concerned, “to attribute a historical value to
. . . the slender links with northwestern Iran and northern Afghanistan ... is a mistake
“because . . .” it could well be the spread of particular objects and, as such, objects that
could circulate more easily quite apart from any real contacts” (691 -692).

Jhukar Culture

This culture is situated in the Indus Valley itself and named after Mughal’s excavations
at Jhukar. It includes the sites of Amri and Chanhu-daro. Chanhu-Daro, was excavated
by Mackay, and a report was published in 1943. Allchin draws attention, on the one
hand, to the complete absence of typically Harappan seals and other characteristically
Harappan urban craft products at this site, and, on the other, to the presence of stamp
seals with parallels in eastern Iran, Afghanistan, and central Asia, as well as to a range
of copper tools and pins with analogues in those regions, but equally “foreign” to the
Harappan world. Allchin (1995, 32) is aware that the “foreign” elements noticed at
Chanhu-Daro have not been evidenced at other sites such as those in Saurashtra. He
also acknowledges that “while it is possible to argue, as some have done, that the metal
artifacts of ‘foreign’ type may be no more than die products of trade,” he nonetheless
holds that the presence of the stamp seals is unequivocal in implication: the absence of
Harappan seals—the symbols of Harappan political and economic power—and their
displacement by these foreign seal types “must indicate that a new power was dominant
at Chanhu-daro and by inference in the middle and lower Indus region” (31).

Archaeological Evidence inside the Subcontinent 227

Jarrige (1983), in contrast, had long since pointed out that there were chronological
problems with correlating the seals and other artifacts with the Indo-Aryans: “The mi¬
gration of these [seminomadic] groups Icoming from central Asia] would sometimes be
traced on maps based on the accidental discovery of certain types of artifacts 5 —princi¬
pally metal objects and seals—which could be stylistically associated with the Hissar III C
complex.” He points out, however, that this complex is now dated to the end of the third
millennium B.C. making it contemporary with the Mature Harappan and not later, as
was previously thought: “Thus most of these finds must be interpreted in the context of
international exchange covering the whole of the Middle East and cannot be interpreted
as reflecting the invasion of pastoralists in the mid-2nd millennium bc” (42).

Jarrige (1973) complains of the tendency of lumping everything not typically Harappan
under the rubric of the Jhukar culture, “a problem which is further complicated when,
by attempting to harmonize the archaeological data with philological arguments, people
have developed the habit of attributing to the Jhukar culture all discoveries amenable of
offering some correlation with the Iranian world and Central Asia” (263). Jarrige goes
on to consider whether there was a disruption of sedentary urban life in the Indus Valley
and a sudden drop in agricultural productivity of that area accompanied by a shift to
seminomadic pastoralism with evidence of warfare—in short, all of the features that would
ideally accompany an intrusion of Indo-Aryan nomads. As the excavator of Pirak, the
only well-preserved second millennium b.c.e. site from the area (which he dates from
1700 to 700 b.c.e.), Jarrige (1985) finds a “town” of some size with “elaborate architec¬
ture” and evidence of a more intense level of irrigation and cultivation than occurred in
tire third millennium b.c.e.: “Just the opposite of that which has been presumed on the
basis of negative evidence” (46). In view of the fact that Pirak is widely accepted as her¬
alding the Indo-Aryans due to the discovery of the horse there, my previous remarks
about Indo-Aryans and urban centers are reinforced. Those wishing to consider Pirak
as evidence of nomad Indo-Aryan pastoralists must address the fact that it was “a town
of some size with elaborate architecture” that increased the agricultural productivity of
the area.

Jarrige’s study of continuity and change concludes that the people living in the Kachi
plain during the second millennium b.c.e. undoubtedly experienced the major economic
transformations of the time yet maintained significant elements of cultural continuity
and conservatism from the early third millennium b.c.e. and earlier. He underscores
the continuity aspect of the area by comparing the ancient ruins of residential buildings
from the excavations at Pirak with the very recent ruins of a house deserted by Hindus
at partition in the same district. The resemblance is striking, while the samples of cook¬
ing pots between the two periods seem almost identical. Regarding the transformations,
he doubts whether every newly attested item in the Kachi archaeological record of the
second millennium b.c.e. could be attributed to an influx of new peoples, “since the
processes . . . are too complex to be attributed to the arrival of invaders who at the same
time would have had to have introduced rice from the Ganges, sorghum from the Ara¬
bian Gulf, and camels and horses from Central Asia” (Jarrige 1983, 56).

The discovery of the horse in Pirak, however, must detain us. The evidence consists
of figurines of horses and horsemen. By the end of Pirak II, the figurines are painted
with trappings, and some are wheeled. This is the first evidence of Equus caballus Linn

228 The Quest for the Origins of Vedic Culture

in South Asia that is accepted by Meadow (1997, 309) from the evidence he has been
able to examine so far. While (Jarrige (1983) is skeptical that “the dream of several
generations of scholars—to find a continuous line of sites from the Eurasian steppes
to the Ganges valley all with ‘typical’ grey ware or ‘steppe-style’ pottery which could
be used to map the movements of Indo-Aryan populations” will ever materialize (63),
he does find this attested presence of the horse significant in this regard. It is only
with the introduction of this animal (allowing that this is the earliest uncontested
evidence for the horse) that Jarrige is prepared to consider “groups related to those
from the Eurasian steppes and Central Asian highlands [beginning] to play an im¬
portant role in the functioning of social and economic systems in the northwestern
part of South Asia” (60).

In view of all that I have discussed regarding the horse in the previous chapter, I will
simply note that Jarrige (1 985) specifically mentions that the existence of the Indo-Ary-
ans has “so far only been deduced on the basis of linguistic evidence” (62; my italics).
Otherwise, “what we see is a dynamic system of multidirectional contacts and ‘influ¬
ences’ extending throughout a vast area from southern Central Asia to the Ganges val¬
ley and continuing from the beginning of the 2nd millennium into the 1st millennium
bc” (62). If the linguistic evidence to which archaeologists defer for the status of the
Indo-Aryans as intruders into South Asia is brought into question, then the horse could
simply become another innovatory item introduced into South Asia through trade. In
the absence of irrefutable linguistic evidence, there is no reason to feel compelled to
believe that the introduction of the horse into the subcontinent is indicative of the in¬
troduction of new peoples any more than the introduction of any other innovatory items
of material culture (such as camels, sorghum, rice, lapis lazuli, or anything else) is rep¬
resentative of new human migratory influxes:

There is evidence for the intensification of subsistence practice, multicropping and the
adoption of new forms of transportation (camel and horse). These changes were made by
the indigenous inhabitants, and were not the result of new people streaming into the re¬
gion. The horse and camel would indicate connections with Central Asia. The cultiva¬
tion of rice would connect with either the Late Harappan in the Ganga-Yamuna region or
Gujarat. (Kenoyer 1995, 227; my italics)

Moreover, if the arrival of a new people can be promoted on the basis of the introduc¬
tion of an essential new item from central Asia, then why should using an inverse logic
denying any such arrival because of the nonintroduction of an essential item from central
Asia not also be given consideration? According to Kenoyer (1995):

In earlier models, the northwestern regions were the source of the so-called movements
of Indo-Aryan speaking peoples. Yet, if there were such movements, why were the mi¬
grants not supplying one of the most important raw materials for bronze production, i.e.
tin? This cannot be answered simply by saying that iron was replacing copper and bronze,
because the prominent use of iron does not occur until much later, in the NBP [North¬
ern Black Polished Ware] period. (230)

The Vedas, after all, speak of and value bronze as well as the horse. The argument
here is why did the bronze-using Indo-Aryan speakers not convey tin into the subcon¬
tinent from the northwest if they were arriving from those areas?

Archaeological Evidence inside the Subcontinent 229

Cemetery H Culture

The Cemetery H excavations, a description of which appeared in a report published by
Vats in 1940, contained pottery that the excavator considered indicated a “Rigvcdic”
pattern of belief. Wheeler (1947) noted:

The intrusive culture, as represented by its pottery, has in origin nothing to do with the
Harappa culture; its ceramic differs from that of the latter both in finish and in decora¬
tion, and its dwellings . . . are notably more roughly constructed than those of Harappa
proper. Its analogues have not yet been identified, and it appears in fact as abruptly as
did its Harappan predecessor. The suggestion has been made [by Childc] very hesitat¬
ingly, that the Cemetery H intruders “may belong to the Aryan invaders.” (81)

He then proceeded to support this suggestion using the types of arguments I outlined
in the previous chapter. This line of interpretation is still occasionally found in cur¬
rency: “The evidence is admittedly slender, but there appears to be a good case for see¬
ing in the Cemetery H culture the presence of an element of foreign intruders who have
dominated the existing population and exploited their craft products, though modified
to suit their own tastes” (Allchin 1 995, 33). 6 Nowadays few South Asian archaeologists
would concur with this opinion, and I shall simply note with Kcnoyer (1991b) that
Cemetery H “may reflect only a change in the focus of settlement organization from that
which was the pattern of the earlier Harappan phase and not cultural discontinuity,
urban decay, invading aliens, or site abandonment, all of which have been suggested in
the past” (56).

Painted Gray Ware Culture

The Painted Gray Ware (PGW) type of pottery was especially promoted by B. B. Lai
(1978) as best representing the Aryan presence. Outside of professional archaeological
circles, it still surfaces to this day in discussions concerning evidence of Aryan intru¬
sions. The PGW is found in quintessential Aryan locales such as Hastinapur, Mathura,
and Kurukshetra, as well as in the SarasvatT valley from where the Aryans are supposed
to have arrived. It also coincides with the earliest discovery of iron in the archaeological
record, and with findings of horse bones (e.g., Hastinapur). Lai’s theory, if viewed from
within the framework that the Aryans must have come into India in the mid-second
millennium B.C.E., seemed persuasive at the time. Objections had already been raised,
however. Chakrabarti (1968), for example, pointed out that sites like Hastinapur con¬
tain evidence of rice use and of the presence ot domestic pig and buffalo alongside the
PGW, which are all features that have been ascribed an eastern origin, as opposed to
the traditional northwestern origin of the Aryans. Moreover, B. K. Thapar (1970, 156)
remarked that had the PGW been symptomatic of the Aryans coming in from the
Northwest, then the same pottery type would be expected to occur in Iran and Afghani¬
stan. It does not, which threw serious doubts on Lai’s thesis. As far as archaeologists
like Shaffer are concerned, “there is no connection between the PGW and the ‘Ary¬
ans’”; it is an indigenous culture. Accordingly, “If PGW has an indigenous South Asian

230 The Quest for the Origins of Vedic Culture

origin it cannot, therefore, represent an intrusive culture with a western origin.” Most
important from the perspective of this chapter, “this conclusion . . . means that we
have no archaeological culture which might represent the Aryan phenomenon” (Shaffer

1986, 232).

In any event, Lai (1997) himself has changed his views, and his recently published
book contains an appendix called “It’s Time to Rethink” (281 -287). Pressed to address
the Indo-Aryan issue because of his preeminent status in Indian archaeology, and due
to the immense controversy in India surrounding the origins of the Indo-Aryans, he
proceeds to analyze all the arguments that are commonly raised against the possibility
that the Indo-Aryans could have been Harappans. He concludes:

Is it not time to rethink about the entire issue? Could the chalcolithic people of Mehrgarh
[seventh millennium b.c.e.J, who in the course- of time evolved into Bronze Age
Harappans, themselves have been the Indo-Aryans? These chalcolithic people had rela¬
tionship with areas now compromising northern Afghanistan, northeastern Iran and
even southern part of central Asia—which area may have been the habitat of the Aryans
prior to the composition of the Rgveda. (287)

Lai’s proposal, of course, must address the chronological issues involved in dating the
Vedic texts. But it does illustrate the tendency among South Asian archaeologists to
emphasize continuity and trade, rather dran innovation due to migrations. In conclu¬
sion, cultural evidence of the Indo-Aryans whether in central Asia or within the sub¬
continent, cannot be readily traced in the archaeological record. What about the Indo-
Aryan speakers themselves? Can they be connected with a specific racial type in the
archaeological record?

Aryans in the Skeletal Record

Initially, Guha, in 1935, had identified four racial elements at Mohenjo-Daro. Kumar,
who examined the cranial remains from the Harappan material in 1973, suggested that
the rugged proto-Nordic elements from these might be indicative of an influx of Indo-
European or Indo-Iranian people, despite acknowledging the difficulty in identifying
racially distinct Aryans (Kumar 1973, 74). His starting premise, however, was that “the
Aryan invasion ... is accepted on the basis of the evidence of the Vedas” (67), by which
he was referring to the racial references outlined in chapter 3.

K. Kennedy (1984), however, who was able to examine all three hundred skeletons
that had been retrieved from the Indus Valley Civilization, found that the ancient
Harappans “are not markedly different in their skeletal biology from the present-day
inhabitants of Northwestern India and Pakistan” (102). He considers any physical varia¬
tions in the skeletal record to be perfectly normal for a metropolitan setting and consis¬
tent with any urban population past or present (103). As far as he is concerned, the
polytypism in the South Asian record represents an “overlap of relatively homogeneous
tribal and outcaste groups and their penetration into villages, dien into urban environ¬
ments of more heterogeneous people.” There is no need to defer to intruding aliens for
any of this: “This dynamic rather than mass migration and invasions of nomadic and
warlike peoples better accounts for the biological constitutions of those earlier urban

Archaeological Evidence inside the Subcontinent 231

populations in the Indus valley.” Here, again, we encounter the same objections raised
repeatedly by South Asian archaeologists: “Of the Aryans, we must defer to literary and
linguistic scholars in whose province lies the determination of the arrival and nature of
the linguistic phenomenon we call the Aryans. . . . But archaeological evidence of Aryan-
speaking peoples is questionable and the skeletal evidence is nil” (104; my italics).

Not only is the skeletal evidence nil, but “if invasions of exotic races had taken place
by Aryan hordes, we should encounter obvious discontinuities in the prehistoric skel¬
etal record that correspond with a period around 1500 BC.” Whatever discontinuities
do occur in the record are either tar too late or far too early (Kennedy 1995, 58). These
discontinuities were taken from a further study undertaken on the skeletal remains in
the Harappan phase “Cemetery R37” (Hemphill et al. 1991). The results of this survey
showed two periods of discontinuities: the first occurs during the period between 6000
and 4500 b.c.e. between the Neolithic and Chalcolithic inhabitants of Mehrgarh, and
the second at some point before 200 b.c.e. (but after 800 b.c.e.), which is visible in the
remains at Sarai Khola (200 b.c.e.). Clearly, neither of these biological discontinuities
corresponds with the commonly accepted period for Indo-Aryan intrusions. The Ary¬
ans have not been located in the skeletal record. It is important to note, at the risk of
repetition, that Kennedy, like almost all South Asian archaeologists, has deferred to the
“literary and linguistic” evidence for the arrival of the Aryans, since archaeology has not
uncovered any trace of them.

Continuity and Innovation

Archaeologists today speak of “integration,” “decentralization,” “localization,” and
“regionalization” to characterize the relationship of the various cultures of Northwest
India/Pakistan with the Mature Harappan period (Shaffer 1992). In the Post-Harappan
period, various regions, such as Jhukar and Cemetery H, arc seen as “springing from
the Mature Harappan but having distinct assemblages localized in their own areas”
(Mughal 1988). Aryans are not a mandatory ingredient in this process:

Although the overall socioeconomic organization changed, continuities in technology,
subsistence practices, settlement organization, and some regional symbols show that the
indigenous population was not displaced by hordes of Indo-Aryan speaking people. . . .

For many years, the “invasions" or “migrations” of these Indo-Aryan-speaking Vedic/
Aryan tribes explained the decline of the Indus civilization and the second rise of urban¬
ization. . . . This was based on simplistic models of culture change and an uncritical read¬
ing of Vedic texts. Current evidence does not support a pre- or proto-historic Indo-Aryan
invasion of southern Asia. . . . Instead, there was an overlap between Late Harappan and
post-Harappan communities . . . with no biological evidence for major new populations.
(Kenoyer 1991a, 371)

The vast majority of the professional archaeologists 1 interviewed in India insisted that
there was no convincing archaeological evidence whatsoever to support any claims of
external Indo-Aryan origins. This is part of a wider trend: archaeologists working out¬
side of South Asia are voicing similar views. Lamberg-Karlovsky (1993) comments on
the extraordinary complexity and considerable debate within the archaeological litera¬
ture on the issue of external versus internal “causal premises” for the origin of the Bactrian

232 The Quest for the Origins of Vedic Culture

Bronze Age: “It must be admitted that within recent years there has been a penchant to
emphasize the indigenous nature of social processes. While vaguely admitting to some
degree of interaction archaeologists have emphasized the autochthonous nature of virtu¬
ally every archaeological district” (34). R. Dyson (1993) remarks, in his discussion of
changes taking place in the field, that the invasion thesis ’’becomes a paradigm of lim¬
ited usefulness” (576). He proposes that “by freeing themselves from this hypothesis
drawn from earlier linguistic studies, archaeologists may now focus their attention on
the archaeological evidence in its own terms” (576). Commenting on the “continuing
lack of agreement over the criteria by which the presence of the Indo-Aryans can be
demonstrated,” he outlines the alternative paradigm taking shape in the archaeology of
the whole region I have been discussing: “The suggestion of an indigenous Indo-Aryan
population going far back into pre-history in Northeastern Iran and nearby Turkmenia
is now taken quite seriously.” With this trend in mind, he finds it interesting that the
discussion between contributors of PossehPs Harappan Civilization “indicated a parallel
trend” (577).

Among Western archaeologists, Jim Shaffer has been the most outspoken critic of
the Aryan invasion theory. According to Shaffer (1984b):

Current archaeological data do not support the existence of an Indo-Aryan or European
invasion into South Asia at any time in the pre- or protohistoric periods. Instead, it is
possible to document archaeologicaily a series of cultural changes reflecting indigenous
cultural development from prehistoric to historic periods. . . . The Indo-Aryan invasion(s)
as an academic concept in 18th- and 19th-century Europe reflected the cultural milieu of
that period. Linguistic data were used to validate the concept that in turn was used to
interpret archaeological and anthropological data. What was theory became unquestioned
fact that was used to interpret and organize all subsequent data. It is time to end the
“linguistic tyranny” that has prescribed interpretative frameworks of pre- and proto-his¬
toric cultural development in South Asia. (88)

Shaffer complains that archaeological data can be artificially constrained to support com¬
monly held beliefs and presents the case of urbanization in the Northwest of the sub¬
continent as an example. As has been noted, it is generally held that the Vedic texts
represent a nonurban pastoral society, in sharp contrast to the cities and towns of the
Indus Valley. Scholars have generally postulated a distinct break of over a millennium
between the urban centers of the Indus Valley and the reemergence of urbanization in
the Gangetic Plain. This hiatus is taken to correspond neady to the period assigned to
the arrival of the supposedly nonurban Aryan nomads sometime in the second millen¬
nium B.C.E.

Shaffer (1993) refers to one set of data that undermines this simplistic portrayal of
an apparent devolution and re-evolution of urbanization, which “has nearly become a
South Asian archaeological axiom” (55). Although there appears to have been a definite
shift in settlements from the Indus Valley proper in late and Post-Harappan periods,
there is a significant increase in die number of sites in Gujarat, and an “explosion”
(300 percent increase) of new settlements in East Punjab to accommodate the transferal
of the population. Shaffer (1995) is insistent that “this shift by Harappan and, perhaps,
other Indus Valley cultural mosaic groups is the only archaeologicaily documented west-to-east
movement of human populations in South Asia before the first half of the first millennium
B.C.” (139; italics in original).

Archaeological Evidence inside the Subcontinent 233

Moreover, although there is a general decrease in the size of tire settlements not all
of these were small and insignificant in comparison with the large, complex structures
of the Mature Harappan period. Data from Bahawalpur, the region in Pakistan most
thoroughly surveyed, suggests an increase in size in the settlements of the Late Harappan
period in comparison to the Harappan period (Shaffer 1993, 57). This is very signifi-
cant: “More surveys have revealed large, post-Harappan settlements in the Indus region
after the major Indus centres were abandoned. . . . Research ... is beginning to dem¬
onstrate that there really is no Dark Age isolating the protohistoric from the historic
period” (Kenoyer 1987, 26). 7

As with the BMAC culture, these data also problematize the notion that the Vedas
know no urban centers and therefore must be Post-Harappan. If the Vedas are, in fact,
silent regarding large settlements, it is not because of a lack of such settlements at the
approximate time and place the hymns are assumed to have been composed (viz., Punjab
in the mid-second millennium b.c.e.) because, as Shaffer has noted, settlements did
not disappear; they simply shifted east. This would have been caused by the hydrologi¬
cal, ecological, and other factors mentioned previously that had struck sites farther west.
Thus, there was “reorganization and expansion” (Kenoyer 1995, 234), but not dissolu¬
tion. And Pirac, we can recall, was a town of some size.

Once it had become established lore that the Indo-Aryans had intruded into the
subcontinent around the middle of the second millennium b.c.e., archaeologists natu¬
rally began looking for innovations in the archaeological record that could be used as
evidence of their arrival. D. K. Chakrabarti (1968) decades ago voiced the by now famil¬
iar complaint among South Asian archaeologists that such an a priori precommitment
blinkered and actually hampered proper examination of the archaeological material in
its own light: “To what extent has this Aryan hypothesis contributed to a better under¬
standing of the relevant Indian archaeological data. 7 In two cases at least. . . the Painted
Grey ware and Ahar cultures this seems to have actually distracted attention from the
basic task of a proper evaluation and analysis of the cultures themselves” (358).

The Aryans have not been pinpointed in the archaeological record with anything
approaching general consensus among archaeologists. B. B. Lai, in his earlier work,
demonstrated the inadequacy of other material cultures that had been identified as Aryan. 8
B. K. Thapar (1970) undertook a similar process of elimination. There is more or less
unanimous agreement among present-day Indian archaeologists that there is no mate¬
rial culture that can be identified with any incoming Aryans. Clearly, any proposals
made in the past could only have appeared attractive, that is to say, less defective than
other alternatives, because their advocates had embarked on their investigations with
the assumption that the Aryans must have entered rhe subcontinent in the Late or Post-
Harappan period and must, therefore, sooner or later correspond to some innovative
archaeological entity. In Chakrabarti’s estimation, all claims of Aryan identification have
been either much too general to be meaningful, 9 positively misleading, drawn from cul¬
tural assemblages sometimes separated by as much as several millennia, or connected
to West Asia, from where not a single feature occurs in the supposedly corresponding
Indian counterparts (Chakrabarti 1977b, 33-34).

Other artifacts, often solitary pieces, such as the shaft-hole ax from a late level of
Mohenjo-Daro, the Rajanpur sword, and the trunnion ax, all of which are considered
to be non-Indian but have wide currency in Afghanistan, eastern Iran, and the Oxus,

234 The Quest for the Origins of Vedic Culture

are also sometimes heralded as proof of an alien Indo-Aryan presence. Heine-Geldern
(1936), an early forerunner in this regard, while admitting that “we may not as yet say
with certainty whether these . . . shapes have been brought to India by trading inter¬
course or by an ethnical migration” (104), nonetheless was inclined to consider that
“they were from a later date than the Indus civilization and possibly belonging to the
Vedic age” (23). Chakrabarti (1977b) finds that they “may more satisfactorily be ex¬
plained as nothing more than what they apparendy are: isolated objects finding their
way in through trade or some other medium of contact, not necessarily any population
movement of historic magnitude” (31). He notes that prior to the artificial boundaries
demarcated by the British, the southern part of the Oxus, eastern Iran, Afghanistan,
and the Northwest of the subcontinent all constituted an area with significant economic
and political interaction throughout the ages—a sphere of activity distinct from the Ira¬
nian heartland to the west and Gangetic India to the east. In such an economic and
geopolitical zone, “any new significant cultural innovation in any one area between the
Oxus and the Indus is likely to spread rapidly to the rest of this total area" (31). As far
as he is concerned, “the archaeological data from the Indus system and the area to its
west. . . . which have been interpreted as different types of diffusion from a vague and
undefined West Asia are no more than the indications of mutual contact between the
geographical components of this interaction sphere” (35).

Chakrabarti also voices the by-now familiar complaint that Indian archaeology is overly
dominated by diffusionism—a method of interpretation that views every innovation in
the archaeological record as evidence of the movements of a particular people. B. Singh
(1995) uses a modern example to better contextualize the potential interpretative ex¬
cesses resulting from this method:

If we count the gradual increase of china in Indian houses during the recent decades, and
sit down to judge as future archaeologists with as mechanical an approach as is evident
among those who equate pottery with people, we would find the entire country suffused
by a new people. The same inference would be drawn from the sudden popularity of
stainless steel followed by aluminum alloy and the like. (137)

Although such interpretations are no longer much in vogue elsewhere, the quest for
elusive Indo-Aryan intruders ensured their survival for much longer in Indian archae¬
ology. Chakrabarti (1997b) believes such constraints have severely crippled the visual¬
ization of “protohistoric growth in inner India in its own terms, without any reference
to the supposed multiple waves of people pouring in from the West” (35). The same
could be said for the archaeological cultures generally associated with the Aryans on
their supposed route from the Caspian Sea to the subcontinent.

Erdosy (1995a), who is prepared to find “some support” for small-scale migrations
associated with the intrusive BMAC elements noted earlier, nonetheless states: “Sev¬
eral cultural traits with good Vedic and Avestan parallels have been found widely dis¬
tributed between the southern Urals, Central Asia and the Indo-Iranian borderlands.
However, even allowing for the uncertain chronology of Central Asian sites, few of these
traits show the northwest-southeast gradient in chronology predicted by our linguistic
models.” Rather, in the manner of other traits commonly associated with the “Aryans”
within South Asia, “they originate in different places at different times and circulate
widely, undoubtedly through the extensive interaction networks built up in the mid-3rd

Archaeological Evidence inside the Subcontinent 235

to early 2nd millennia B.C.” The main point is that “it is impossible, thus, to regard
the widespread distribution of certain beliefs and rituals, which came to be adopted by
Indo-Iranian speakers, as evidence of population movements” (12).

Not surprisingly, this basic line of interpretation is favored by Indigenous Aryanists.
Once the underlying equation that diachronic commonalties of the archaeological record
in adjacent geographic areas equaling the physical movement of peoples has been brought
into question, there is little to compel Indigenous Aryanists to accept the theory of the
migrations of the Indo-Aryans from an archaeological perspective. S. P. Gupta’s two-
volume survey (1979) on the archaeology of central Asia and the Indian borderlands,
for example, also concludes that although artifacts from central Asia do occur sporadi¬
cally in the Indian borderlands and the Indus basin, this phenomenon can “best be
interpreted in terms of exchanges” (318). He points out that none of the central Asian
protohistoric cultures reached the Indus region in totality as would have been expected
had large groups of Indo-Aryans been on the move. Along the same lines, a doctoral
dissertation focused exclusively on ethnic movements in the second millennium b.c.e.
between the Caspian Sea and the Indus basin concludes that “there is no case of the
Aryan migration from West Asia to India in the first part of the second millennium
b.c. If there is any case, the case is that of economic ties” (Kesarwani 1982, 312).

A few points need to be made from the perspective of Aryan migrations, however.
The fact that no satisfactorily consistent archaeological culture can be found stretching
across central Asia or the steppes does not necessarily dispel the possibility that large
numbers of people could nonetheless have been on the move. Anthony (1986) draws
attention to an example from the historic period: the Helvetii. The migration of these
people which was recorded by Caesar in some detail, involved the movement of a popu¬
lation mass said to have numbered 360,000 initially and found to number 110,000 in
Caesar’s military census of the defeated remnant. According to Anthony (1986), “Even
assuming a certain amount of exaggeration, this was a very substantial population
movement and one that current Western archaeology theory would neither predict nor
explain. Caesar’s account makes it clear that this was not a unique event” (300). The
same can be said for the horse-riding Huns: “There is yet no answer to the question of
what happened to the mortal remains of these fearful conquerors and their strange
mounts. Hun domination was short-lived and if the dead were cremated and the horses’
bodies not put into the graves, the likelihood of finding their bones is necessarily lim¬
ited” (Sinor 1990, 203). Mallory (1989, 166) underscores the same point by noting the
absence of archaeological evidence to substantiate the introduction of Gaelic into Scot¬
land in the fifth to sixth centuries A.D.

Elsewhere, Mallory (1997) complains that the “argument of archaeological continuity
could probably be supported for every IE-speaking region of Eurasia where any archaeolo¬
gist can effortlessly pen such statements as ‘while there may be some evidence for the
diffusion of ideas, there is no evidence for the diffusion of population movement’” (104).
India is not the only Indo-European-speaking area that has not revealed any archaeologi¬
cal traces of immigration. But the Indo-Europeans must have come from somewhere. If
archaeology cannot confirm the trail of the Indo-Aryans, it cannot deny it either:

The critical point is that language and ethnic shift can take place without radical change
in the material particulars of life and with an amount of change in the gene pool so small

236 The Quest for the Origins of Vedic Culture

as to be for all practical purposes undetectable. We should not replace the fallacy of as¬
signing all significant culture change to migration with the fallacy of thinking that lan¬
guage shift and the spread of new ethnic self-identification occur only with major or radi¬
cal cultural transformations. (Ehret 1988, 565)

However, we must bear in mind that such arguments work both ways. They cannot be
introduced to rebut the objections of those archaeologists working in Soudt Asia who
see no evidence of Indo-European immigrations into the subcontinent, and then be
denied to anyone choosing to argue for an emmigration of Indo-Europeans from the
same subcontinent despite the lack of archaeological evidence. We must grant, how¬
ever, that there is at least a series of archaeological cultures that can be traced approach¬
ing the Indian subcontinent, even if discontinuous, which does not seem to be the case
for any hypothetical east-to-west emigration.

Conclusion

Postmodernism has impacted archaeology under the rubric of post-processualism, which
holds that every reading or decoding of a text, including an archaeological text, is an¬
other encoding, since all truth is subjective. As Trigger (1995) notes, some scholars
maintain that all archaeological interpretations are subjective constructions that are
constrained scarcely, if at all, by archaeological data:

The claim by archaeologists to be able to falsify interpretations on the basis of new evi¬
dence or by means of new techniques of analysis is dismissed by extreme relativists as an
untenable manifestation of elitism and intellectual hegemony, which must be resisted
with counterclaims that all interpretations of the past are subjective and hence there is no
way to demonstrate that the insights of a professional archaeologist are necessarily supe¬
rior to those of anyone else. (276).

The same, of course, could be said about linguistics. Trigger rejects such relativity and
believes that as disciplines such as archaeology develop, they acquire a larger database
and new methodologies that act as constraints on the imagination and excesses of scholarly
interpretation. In the South Asian case, there seems to be a growing disjunct between
philology and linguistics, on one hand, and archaeology, on the other, and this dichotomy
seems to only increase as the database grows larger.

There is little to conclude in this chapter that has not been stated or paralleled in the
previous chapter. As with the central Asian material, innovative items in the archaeo¬
logical record can be seen either as items exchanged by trade and other forms of inter¬
action, perhaps between different economic groups such as pastoralists and agricultur¬
alists, as has always been the case on the subcontinent to this day. This interpretation
seems favored by a significant number of South Asian archaeologists. Or, such items
can be read as possible remanants of a new language group—the Indo-Aryan speakers—
as others hold. Or there may well have been a linguistic intrusion which did not leave
any distinct cultural remnants that could ever surface in the archaeological record. In¬
terpreting the archaeological evidence from within the subcontinent will, to a great ex¬
tent, reflect the attitude one holds toward the linguistic evidence outlined in the previ¬
ous chapters.

Archaeological Evidence inside the Subcontinent 237

Those who find die conclusions drawn from the linguistic evidence compelling will
consider it legitimate to identify some of the archaeological cultures outlined in this
chapter as evidence of an intruding linguistic group. R. S. Sharma (1998), in a barely
veiled reference to Hindu nationalism, finds that “the commitment to cultural continu¬
ity .. . reminds us of the eternal sanatana dharma [eternal religion] propagated in present-
day India” (95). I will discuss the Indian social context of all this in chapter 13. Sergent
finds antimigrationist positions “obscurantist” and points out that India was invaded
nine times in one millennium by Achemenides, Macedonians, Bactrians, Greeks, Sakas,
Kusans, Sassanides, Yuezi, and Hephtalite Huns, just in antiquity. And then, of course
came the Turks, Mongols, Afghans, Portugese, French, and British. How, then, he
wonders, can one realistically deny invasions or migrations? What needs to be men¬
tioned here, however, is that these migrations are not quite adequate comparisons with
any migrations that might be postulated for the Indo-Aryans because none of these groups
eradicated the preexisting languages on the subcontinent as the Indo-Aryans are assumed
to have done. They did, however, act as adstrata and superstrata, and consequently might
have added loanwords or other linguistic features. I discussed the possible parallels to
this in chapter 5.

Those finding the linguistic evidence more inconclusive will, in the absence of com¬
pelling contrary proof from archaeology, likely be more inclined to at least consider the
possibility of an indigenous Indo-Aryan language group in the Indian subcontinent with
all that this entails. Those archaeologists who consider that much of the linguistic evi¬
dence supporting such migrations can be called into question are not likely to feel con¬
strained to interpret the archaeological record under these parameters unless the archaeo¬
logical evidence itself calls for such an interpretation in its own right. But this does not
legitimate proclaiming that the theory of Aryan migrations has been disproved as some
Indigenists feel entitled to do. Far from it: any and all of the archaeological cultures
examined in this or the previous chapter could have corresponded to an intruding Indo-
Aryan ethnic group. An Indigenous Aryan position can only hope to coexist with the
Migrationist’s position, not displace it, at least on the grounds of the presently available
data.

But even this position has a corollary in addition to the fact that an Indigenist posi¬
tion requires that all the Indo-Europeans came from the Northwest of the subcontinent
and its environs—for which no compelling evidence has yet been produced. If we allow,
for arguments sake, that much of the linguistic and archaeological evidence is ambigu¬
ous, there is still a massive chronological obstacle to an Indigenous Aryan position.
How far back could Indo-Aryan have existed on the subcontinent? Since all eternity? I
conclude this chapter as I have several preceding ones: before we can explore any pos¬
sibility of a hoary Indo-Aryan language any further, we need to first direct our attention
to the dating of the Veda, which is the topic of the next chapter.

The Date of the Veda

Everything hinges on the date of the Vedas. Indispensable support for the Indigenous
position would result if the possibility of a much greater antiquity for the Vedic corpus
could be convincingly demonstrated. Indeed, as I have noted in previous chapters, the
Indigenous case actually loses plausibility unless such antiquity can be demonstrated.
On the other hand, if, as some Indigenous Aryanists would have it, the Rgveda is a
thousand or more years older than the date of 1500 b.c.e. presently assigned to it by
most Indologists, a variety of issues will be affected. Since the Vedic horizons are solidly
situated in the Northwest of the subcontinent, a much stronger case could be made
supporting an Indo-Aryan presence in, or coexistence with, the Indus Valley Civiliza-
tion, which shares much of the same horizons in approximately 2500-1900 b.c.e. The
whole horse argument becomes less compelling whilst those promoting the SarasvatT
evidence become vindicated. In addition, there would be very strong grounds for would-
be decipherers to approach the script as containing an Indo-Aryan language.

If the Rgveda is at least a millennium older than its commonly accepted date, then
the possibility of Dravidian and/or Munda and/or unknown linguistic influences on
Vedic Sanskrit being the result of the speakers of these languages intruding on an Indo-
Aryan-speaking area after the other languages had already left, as opposed to vice versa,
becomes a much more serious consideration. Moreover, the relationship between Vedic
and Proto-Indo-European would need to be reconsidered. Any proposal associating the
overland trajectory of the Indo-Aryans with the Andronovo culture, a southern Iranian
route, or any Post-Harappan culture in the subcontinent loses value. For these and other
reasons, a much older date for the Veda is foundational to the Indigenous position and
is promoted almost universally by those adhering to this point of view (whether aware
of all these implications or not). If by contrast, the oldest strata of the Rgveda cannot be

238

The Date of the Veda 239

far removed from the conventionally accepted date of 1200 or 1500 b.c.e. (or, with
Witzel, even 1900 b.c.e.), then the Indigenous case looses cogency. Not far before the
oldest strata of these texts is the joint Indo-Iranian period, which, in an Indigenist sce¬
nario, would have to be correlated with the Indus Valley Civilization proper if the con¬
ventional dates for the Rgveda are accepted; prior to that is the late proto-Indo-Euro¬
pean period, which would have to be contemporaneous with the pre- and early Harappan
period. Neither of these propositions would attract the attention of serious scholars in
the field for some of the reasons already outlined at length (apart from anything else, no
one has attempted to reconstructed the Indo-lranians or late Indo-Europeans as urban
dwellers living in such a sophisticated civilization). Indigenists, then, must demonstrate
that the Rgveda could be at least a thousand, fifteen hundred, or even two thousand
years older than has been generally accepted.

Dating Proto-Indo-European

Obviously, any dating for Indo-Aryan must be posterior to the terminus post quo date
established for the undivided Proto-Indo-European language prior to its disintegration
into the various cognate languages. The terminus ante quern date can be stated with
some security: Anatolian is attested in Akkadian trading documents of about 1900 b.c.e.
and is subsequently followed by the emergence of Hittite, Palaic, and Luwian in written
texts. Indo-Aryan is attested in the Mitanni kingdom by 1600-1500 b.c.e., while Linear
B Greek can be dated to 1300 b.c.e. So it is safe to assume, and is universally accepted,
that by about 2000 b.c.e., Proto-Indo-European was already differentiated, and that Indo-
Aryan was already a linguistic entity distinct from the Iranian speakers by 1600 b.c.e.
at the latest.

The terminus a quo is much more problematic. All commonly accepted dates have
been based on archaeological finds—no linguistic means attempting to document the rate
of language change has been proposed since the rejection of glottochronology as valid.
Glottochronology was a method introduced by Swadesh, whereby a word list of one hundred
common vocabulary items was drawn up (this was later increased to two hundred). The
idea was that time periods could be assigned to the intervals between different stages of a
language, or between cognate dialects and their mother tongue, depending on what per¬
centage of these lexical items had been preserved in the stages or dialects under compari¬
son and what percentage had been altered or changed. For example, by calculating the
amount of change in these one or two hundred items between, say, the English of Chaucer
and that of Shakespeare and correlating that percentage of change with the time known
to have transpired between these two stages of English (and comparing this with similar
calculations from other known stages of historical languages), an attempt can be made to
establish a rough overall formula such as y percentage erf language change approximately
equals x period of time. This formula might then theoretically be applicable in calculating
the rate of change evidenced between unknown reconstructed pre- or protohistoric lan¬
guages and later known stages of these languages and hence provide approximate dates
for a language such as Proto-Indo-European.

The idea is actually ingenious and seemed very promising until it was pointed out
that languages do not change at standard rates. Lithuanian, for example, preserves very

240 The Quest for the Origin of Vedic Culture

archaic Indo-European features to this very day. The method was accordingly discarded
as unreliable by most linguists. Recently, however, Starostin (1 999) has attempted to
revive the method. He maintains that Swadesh’s method should not be jettisoned but
improved in a variety of ways (such as using roots, rather than words, as comparisons
since roots tend to have better retention rates). Pejros (1999), too, sees value in fine-
tuning the method of gluttochronology (with more carefully formulated word lists).

Because, in the absence of dateable inscriptions, there are no other ways accessible
to linguistics that can be used to date languages, Indo-Europeanists defer to archaeol¬
ogy. However, the archaeological method of dating languages is predicated on the clues
offered by linguistic paleontology, so the data invoked to date the protolanguage assume
some validity to this method. Since at least the time of Otto Schrader (1890), scholars
have pointed out that there are cognate words in various languages for copper (Sanskrit
ayas), but none for bronze or iron. Proto-Indo-European is accordingly deduced to have
preceded the Bronze Age and Iron Age, but to have corresponded to the Copper Age.
This evidence entails accepting a date for the differentiation of the Indo-European lan¬
guages earlier than the attestation of bronze—4500 b.c.e. for arsenic-copper alloys, and
at least 3500 b.c.e. for copper tin bronzes (Mallory 1975, 32)—and later than the attes¬
tation of copper, which is known in Europe since 5500 b.c.e. (Mallory 1975, 31). Thus
this method will lead us to a date between 5500 and 3500 b.c.e. for an undivided (and
therefore pre-Indo-Aryan) Proto-Indo-European

Historically, the domestication of the horse has been another often-cited chronologi¬
cal indicator on the grounds outlined in chapter 6. Anthony, Telegin, and Brown (1991)
have proposed a date as early as 4000 b.c.e. for domestication based on the discovery
of bit wear on horse premolars discovered in the Sredni Stog site of Dereivka. How¬
ever, since, as has been discussed, many linguists have challenged the assertion that the
IE’s necessarily knew a domesticated horse rather than some other type of equid, the
reliability of this data can be questioned. In a more recent article, Anthony (1995a)
relies more heavily on the evidence of the wheel to make his case. He argues that the
PIE’s were undoubtedly familiar with wheeled vehicles, since at least six different terms
have been reconstructed: three for ‘wheel’, one for ‘axle’, one for ‘thill’, and a verbal
root meaning to ‘convey in a vehicle’ (there is no shared root for ‘spoke’, which could
suggest that this development was a later technology). Wheeled technology appears be¬
tween 3300 and 3100 b.c.e. in four different media in Europe and the Near East. 1
Anthony (1994) accordingly concludes that this evidence “requires a dispersal no ear¬
lier than about 3300 b.c. (192). If this reasoning is valid, it sets some kind of a termi¬
nus a quo for Indo-Aryan as a differentiated speech community.

Some major assumptions are embedded here that one must accept before agreeing
with this conclusion. First, one must be comfortable with the techniques and efficacy of
archaeological dating. Next, one must allow some validity to the method of linguistic
paleontology which, as we have discussed, has been severely criticized by linguists. Ob¬
jections can be raised against almost any diagnostic piece of Indo-European culture that
might be correlatable with the archaeological record. Coleman (1998), for example, who
is sympathetic to Renfrew’s critique of this method, notes that there are four different
roots for the wheel in the Indo-European languages. This suggests to him that “it looks
as if ‘wheel’ was not in the proto-lexicon and the various words for it were created in¬
dependently after the dispersal, in some areas no doubt by loan-translation from adja-

The Date of the Veda 241

cent Indo-European dialects/languages” (451). Along similar lines, D’iakonov (1985)
states that “some processes in which rotating was required were known to mankind
since Palaeolithic times, and we do not necessarily have to associate them with the wheel;
and it has yet to be clarified if the terms for ‘wheel, chariot’ were not used in an earlier
period for ‘potter’s wheel.’” (11 3) I have noted how Lehmann and other linguists have
raised parallel arguments about the horse evidence.

A further assumption one must accept is that the first occurrence of an artifact in
the archaeological record is indicative of its actual date and locus of invention. 2 Anyone
viewing this as an act of faith is likely to remain unconvinced by dates based on the
conjunction of archaeology with linguistic paleontology. I have discussed how in the
Indian context, at least, the spoked wheel was undoubtedly present for a full millen¬
nium or so before surfacing archaeologically, and probably much longer. Be all this as
it may, (and most Indo-Europeanists are not unaware of these problems) when other
factors known to linguistic paleontology and archaeology such as ‘plows’, ‘milk’, ‘wool’,
and ‘silver’ arc factored in along the same lines—most scholars have resigned themselves
to a bracket of about 4500-2500 b.c.e. for the period of the dispersal.

There have been a number ot notable exceptions. As noted, Renfrew has scant regard
for linguistic paleontology and proposed situating the beginning of the Indo-European
dispersal at around 6000 b.c.e. (in order to coincide with the evidence for the spread
of farming). Other scholars have objected. For example, Crossland (1988, 453) points
out that such a time frame would require supposing that the first Greek and Hittite
texts had been written four thousand years after the initial breakup of the Indo-European
languages, and the first Lithuanian and Albanian texts nearly seventy-five hundred years
thereafter. As far as he is concerned, on this time scale one would expect far greater
linguistic diversity in Indo-European than is the case. He draws attention to the diver¬
sity among the Romance languages in somewhat less than two thousand years from
their point of origin as a comparable parallel.

Whether or not the diversity among the Romance languages is at all comparable to
the degree of diversity between Greek and Hittite is debatable, and Renfrew (1988) is
unswayed by all such arguments and challenges his critics to “substantiate more se¬
curely their ‘hunch’ that the proposed time scale is too long”. He notes that, after all,
“all agree that the supposedly regular divergence rates proposed by the practitioners of
glottochronology are to be rejected” (463). Renfrew (1990) has continued to defend this
position adamantly: “Many linguists have commented that these proposed dates of sepa¬
ration are ‘too early,' but how ... do they know this, or judge this?” (19). As will be
discussed later, there are no convincing criteria to determine the conditions in which
conservatism might prevail in a language and over what time period. Earlier, Crossland
(1972) himself had made the following comment:

It is disappointing to have to say that at present there seems to be no hope of estimating
objectively and with a useful degree of precision how long an originally homogeneous
Indo-European language would have taken to develop into derivative groups or languages
which diverged as much as Greek, Sanskrit and Hittite did when the earliest texts in
them were composed. Some linguists seem to think that they can make intuitive judge¬
ments about the minimum time which a particular phonetic or other change in a lan¬
guage would have taken. But the results of intuition when applied to estimating the mini¬
mum time in which a group of cognate languages or dialects would have differentiated to

242 The Quest for the Origin of Vedic Culture

an observed extent vary so much that no useful deductions can be made from them. . . .

I sympathize with archaeologists and other prehistorians who are not primarily linguists
over this. Linguists are unable to provide the information which would be most useful.
(46-47)

Other archaeologists have also argued that the whole reason scholars have failed to
identify the Indo-Europeans or their homeland is because they are looking in a tempo¬
ral bracket that is far too late: some archaeologists believe that the dispersal of the Proto-
Indo-Europeans took place in the much earlier Mesolithic or even Paleolithic (Thomas
1982, 84). Linguists have also argued for a far greater antiquity. Dolgopolsky (1990-
93), in contradistinction to Coleman, unambiguously rejects a post-4500 b.c.e. date,
since “Mallory’s dating, which presupposes that Proto-Anatolian, Proto-Indo-Iranian,
Greek and other descendant languages could have diverged from each other for a mere
2000 years, is absolutely inconceivable” (239). While well aware of the inadequacy of
the glottochronological medaod, Dolgopolsky nonetheless tries to illustrate his argument
by pointing out that all the Germanic languages, over a comparable time depth of two
thousand years, maintained cognates for the same Germanic word for the term ‘hand’
(German hand; Icelandic hond, etc.). Likewise, all the Slavic languages maintain cog¬
nates for their word for hand, *roka, as do the Romance languages for manu. In con¬
trast, over the same time period of 2000 years, the Indo-European languages have not
maintained cognates for one proot-Indo-European term for hand, but developed five
distinct terms for dais basic item of vocabulary. This is conspicuous since body parts are
often the most change-resistant items in a vocabulary. In other words, “if the degree of
closeness between proto-Germanic, proto-Balto-Slavic, proto-Indo-Iranian and other
daughter languages of Indo-European were comparable to that of the modern Germanic
languages, we should expect to find the same word for ‘hand’ in all descendent proto¬
languages of the Indo-European family, i.e., no cases of replacement. The reality is quite
different: we find no less than five cases of replacement” (239). According to Dolgopolsky,
a 4500 b.c.e. time frame is “utterly unrealistic.”

Dolgopolsky, believes the terms for material culture commonly used to date Proto-
Indo-European, such as copper, horse, and wheeled vehicles, cannot be reconstructed
in the Anatolian languages and therefore belong to a later, post-Hittite Indo-European
and not Proto-Indo-European. Proto-Indo-European, accordingly, did not know all these
items, since it was much earlier. He reiterates the arguments outlined previously, sug¬
gesting that the horse known to the Proto-Indo-European’s was not the domesticated
but the wild variety, and claims that the word *ayes did not originally refer to copper
but to metal in general, and it may then later have been transferred to copper in some
countries when this metal entered common usage: “Hence none of these words can
serve as evidence for dating Proto-IE” (Dolgopolsky 1990-93, 241). As noted earlier,
these same arguments have been raised by several linguistics (Coleman 1988; D’iakonov
1988). Dixon (1997) voices similar misgivings:

What has always filled me with wonder is the assurance with which many historical lin¬
guists assign a date to their reconstructed proto-language. . . . We are told that proto-Indo-
European was spoken about 6,000 years ago. What is know with a fair degree of certainty
is the time between proto-1 ndo-Aryan and the modern Indo-Aryan languages—something
in the order of .3,000 years. But how can anyone tell that the development from proto-
Indo-European to proto-Indo-Aryan took another 3,000 years? . . . Languages are known

The Date of the Veda 243

to change at different rates. There is no way of knowing how long it ook to go from the
presumed homogeneity of proto-Indo-European to the linguistic diversity of proto-Indo-
Iranian, proto-Celtic, proto-Germanic, etc. The changes could have been rapid or slow.

We simply don’t know. . . .Why couldn’t proto-Indo-European have been spoken about
10,500 years ago? . . . The received opinion of a date of around 6000 BP for proto-Indo-
European ... is an ingrained one. 1 have found this a difficult matter to get specialists to
even discuss. Yet it does seem to be a house of cards. (47-49)

Zimmer (1988), who prefers a relatively much later date, states: “It must be stressed,
and cannot be said often enough, that whatever date is given for ‘PIE,’ it is necessarily no
more than pure speculation” (372). Zimmer does not mince his words: “Every attempt,
then, to give absolute dates for ‘Proto-Indo-European’ (or dates for alleged different stages
of‘PIE’) is either based on the speculative identification of an archaeological culture with
the speakers of the ‘language of the PIE’s’ (e.g. Gimbutas, Renfrew, Mallory) or on what
may be called ‘intelligent guesses,’ deliberations of probability and feelings of appropriate¬
ness (e.g. Meid, Gamkrelidze-Ivanov)” (372). This results in problems all around: “The
first type of proposal is usually contested by fellow archaeologists and doubted by lin¬
guists, the second, being purely subjective because objective arguments simply do not exist,
is bound to remain noncommittal. As is easily to be seen, many dates of both types have
found their way to an often for too skeptical public” (372). Accordingly, “It is therefore
historically irresponsible for the linguist to speak of‘Proto-Indo-European’ in the 4th millen¬
nium, and linguistically meaningless for the archaeologist to argue about ‘Proto-Indo-Europeans’
living somewhere before ca 2500 b.c.” (374-375; italics in original). In short, these Indo-
European dating conjectures serve as a backdrop to the chronology of the lndo-Aryans.

Dating the Veda

It is from widtin this framework that one must negotiate the date of the Rgveda, which
is the oldest record of the lndo-Aryans apart from the few scanty words revealed in the
Mitanni documents. An exceedingly prominent area of contention among almost all
scholars who have argued for the autochthony of the Aryan speakers in India is the late,
and perceived arbitrary, date that has been assigned to the Vedic texts by most Western
scholars. Such consensus was not always the case. Almost a century ago, Winternitz
([1907] 1962) was refreshingly forthright about the lack of agreement regarding even
the approximate date of the Veda: “It is a fact, and a fact which it is truly painful to
admit, that the opinions of the best scholars differ, not to the extent of centuries, but to
the extent of thousands of years, with regard to the age of the Rg Veda. Some lay down
the year 1000 b.c. as the earliest limit for the Rg Vedic hymns, while others consider
them to have originated between 3000 and 2500 b.c.” (253). Despite such differences
of opinion in this matter, evenmally, communi consensu, the Indological community
settled on 1200 b.c.e. as the probable date for the compilation of the Rgveda-a date
that has remained standard to this day. As opponents never tire of pointing out, “it was
Max Muller who put forth the hypothesis . . . that the Rgveda began to be composed in
1200 b.c.” (Varma 1984, 2).

Muller based his calculations on information he gleaned partly from the Kathasaritsagara,
a collection of stories written in the twelfth century c.E. by Somadeva. In one of these

244 The Quest for the Origin of Vedic Culture

stories, we find a Katyayana Vararuchi, who was reported to have eventually become a
minister in the court of King Nanda. Since, in the Puranas, Nanda is the predecessor of
the Mauryas, Muller assigned him a date in the second half of the fourth century B.C.E.,
shortly before the accepted date for this dynasty. 5 In brief, Muller felt he now had a
reasonably secure date for Katyayana Vararuci. His next step was to correlate this
Katyayana with a Katyayana who was said to have authored a variety of sutras. 4 Since
other sutras were both anterior and posterior to this latter Katyayana, to whom he had
assigned a date in the fourth century b.c.e., Muller ([1859] 1968) decided that “as an
experiment, therefore, though as no more than an experiment, we propose to fix the
years 600 and 200 b.c. as the limits of that age during which the Brahmanic literature
was carried on in die strange style of Sutras” (218).

Preceding the sutras are the Brahmana portions of the Vedic texts (since the latter
are presupposed by the former). Regarding these, Muller ([1859] 1968) considered that
“it would seem impossible to bring the whole within a shorter space dran 200 years. Of
course this is merely conjectural” (395). Conjectural or nett, the Brahmanas, in Muller’s
schema, are consequently assigned a date from 800 to 600 b.c.e., “although it is more
likely that hereafter these limits will have to be extended” (406). Older still than the
Brahmanas are the Mantras, which, in turn, are anterior to the Chandas so, since he
seemed to be on a roll with these concise 200-year brackets, Muller felt that “if we as¬
sign but 200 years to the Mantra period, from 800 to 1000 B.c., and an equal number
to the Chandas period, from 1000 to 1200 b.c., we can do so only under the supposi¬
tion that during the early periods of history the growth of the human mind was more
luxuriant than in later times" (525). As Winternitz ([19071 1962) points out, “it is at
the fixing on these purely arbitrary 1 dates that the untenable part of Max Muller’s calcu¬
lations begins” (255).

Reaction to Muller’s perfectly synchronized, two-hundred-year periods for the develop¬
ment of these different genres of literature was not slow in coming. Goldstiicker ([I860]

1965) objected that “neither is there a single reason to account for his allotting 200 years
to the first of his periods, nor for his doubling this amount of time in the case of the Sutra
period” (80). He points out that, ultimately, “rite whole foundation of Muller’s date rests
on the authority of Somadeva . . . [who) narrated his tales in the twelfth century after Christ
[and] would not be a little surprised to learn that ‘a European point of view’ raises a ‘ghost
story' of his to the dignity of an historical document” (91). Barthelemy Saint-Hilaire (1861)
remonstrated that “Mr. Max Muller would have done well not to have fixed things so
precisely, and not to have circumscribed things so neatly” (54; my translation). H. H.
Wilson (1860) joined in the cacophony of objections to Muller’s methodology:

We must confess that we are disposed to look upon this limit [two hundred years for the
Brahmanas] as much too brief for the establishment of an elaborate ritual, for the appro¬
priation of all the spiritual authority by the Brahmans, for the distinctions of races or the
institutions of caste, and for the mysticism and speculation of the Aranyakas or Upanishads:
a period of five centuries would not seem to be too protracted for such a complete remod¬
elling of the primitive system and its wide dissemination through all those parts of India
where the Brahmans have spread. (376)

Buhler (1894), utilizing evidence from ancient Jain and Buddhist sources, found it in¬
conceivable that “the ancient Indians raced through the so-called Chandas, Mantra and

The Date of the Veda 245

Brahmana periods at a furiously fast pace” (246). Jacobi (1884) constructs a scenario to
give a better conceptual image of the mechanics and time span involved in oral trans¬
mission in the days long before computer technology and electronic media:

It is easy to see that this estimate [i.e., two hundred years] is far below the minimum of
the possible period, during which in India a department of literature could take its rise,
reach perfection, become obsolete and die out, to give place finally to a thoroughly new
departure. For a Brahmana, for example, could only be widely spread by being learned
by heart by a gradually extending circle of Brahmanas, and with the size of the country
this would certainly demand a long time. Every man, who learned such a work, became,
so to say, a copy of it. . . . But several of such works must successively take the place of
their predecessors, before the entire class of works in question becomes obsolete. I main¬
tain that a minimum of a thousand years must rather be taken for such a process, which
in the conditions that prevailed in ancient India was of necessity a very slow one, espe¬
cially when we take into consideration that in historical times die literature of the classi¬
cal period remained for more than a thousand years unaltered. (158)

Each of the periods between the different genres of literature would have required at
least a millennium to develop, spread, and become obsolete for Jacobi, who, as we will
see, had argued for a fourth millennium date for the Rgveda. (Of course, the assign¬
ment of one thousand years to these periods is as arbitrary as Muller’s two hundred
years.) Winternitz (1907), too, felt that since “all the external evidence fails, we are
compelled to rely on the evidence out of the history of Indian literature itself, for the age
of the Veda. . . . We cannot, however, explain the development of the whole of this
great: literature, if we assume as lare a date as round about 1200 or 1500 b.c. as its
starting point. We shall probably have to date the beginning of this development about
2000 or 2500 b.c.” (310; italics in the original).

Max Muller (1892), who hastily acknowledged that he had only considered his date
for the Veda a terminus ad quern, completely submitted to his detractors; “I need hardly
say that I agree with almost every word of my critics. I have repeatedly dwelt on the
hypothetical character of the dates. . . . All I have claimed for them has been that they
are minimum dates . . . Like most Sanskrit scholars, I feel that 200 years ... are scarcely
sufficient to account for the growth of the poetry and religion ascribed to the Khandas
period” (xiv-xv). A few years later, at the end of his long and productive life, he again
acknowledged the complete arbitrariness of his previous calculations: “Whether the Vedic
hymns were composed 1000, or 1500, or 2000, or 3000 years b.c., no power on earth
will ever determine” (Muller 1891,91). Elsewhere, Midler (1897, 87) was quite happy
to consider a date of 3000 b.c.e. based on Sayce’s discovery of two Babylonian ideo¬
graphs—cloth + vegetable fiber (which Sayce believed was cotton)—that had to be pro¬
nounced ‘sindhu. This suggested that the Babylonians knew of the river Sindhu and,
by extension, since he considered this word to be Sanskrit, the Indo-Aryan-speaking
people, in 3000 b.c.e.

However, despite Muller’s willing retraction of his hasty attempt at chronology:

It became a habit already censured by W. D. Whitney, to say that Max Muller had proved
1200-1000 B.c. as the date of the Rg Veda. It was only timidly that a few scholars, like
L. von Schroeder ventured to go as far back as 1 500 or even 2000 b.c. And when all at
once, H. Jacobi attempted to date Vedic literature back to the third millenary B.c. on the
grounds of astrological calculations, scholats Taised a great outcry at such heretical proce-

246 The Quest for the Origin of Vedic Culture

dure. . . . Strange to say it has been quite forgotten on what a precarious footing stood
the “opinion prevailing hitherto,” which was so zealously defended. (Winternitz [1907]
1962 256)

Whitney ([1874] 1987) had made a point of mentioning that Muller himself had made
no pretensions that his dates had “in any essential manner contributed to the final
settlement of the question.” But his concern is that Muller “is in danger of being mis¬
understood as doing so; we have already more than once seen it stated that ‘Muller has
ascertained the date of the Vedas to be 1200-1000 b.c.’” (78). Winternitz (1907), too,
hastened to note that “Max Muller himself did not really wish to say more than that
such an interval at least must be assumed. . . . He always considered his date of 1200-
1000 B.c. only as a terminus ad quern” (293).

These exchanges have been presented in some detail not merely on account of their
historical interest: Mtiller’s initial calculation (albeit no longer for the same reasons) is
still the cornerstone of Vedic dating. As Varma (1984) states, “it is amazing to note that
all the supporters of the date 1200-1000 b.c. for the Veda very conveniently ignore the
caution which Max Muller had initially observed” (6). Indeed, we find present-day
scholars stating that “Max Muller’s chronological estimate, though not devoid of weak
points, has . . . often been more or less tacitly regarded as nearest the mark. ... As far as
the Rgveda is concerned [his] computation is not unreasonable” (Gonda 1975, 22).

To be fair, there are more substantial reasons that have been brought forward to
support Muller’s initial formulations. There are philological clues that can be connected
with archaeology, although these, too, can be brought into question. A summary of
some of the issues raised in the previous chapter is relevant at this point in connection
with their usefulness for dating. The argument that tire Rgveda knows no urban centers
and therefore postdates the Indus Valley Civilization is problematic since archaeology
has revealed large late and Post-Harappan urban settlements exactly where and when
the Aryans are supposed to have been entering. Besides, the Aryans either would have
passed through the BMAC if they entered via the northern route or were the authors
of that urban culture (according to a number of scholars) and established urban centers
along a southern route, if we are to follow Sarianidi. So the Rgveda is not silent about
urban centers due to the Indo-Aryan’s ignorance of them. Moreover, several scholars
have proposed that nomadic and urban cultures must always have, by necessity, coexisted.

Along the same lines, it has been argued that the absence of horse bones and of the
chariot does not have to be synonymous with absence of the Indo-Aryans themselves.
Indeed, the chariot is not attested archaeologically until a full millennium after it was
indisputably present in South Asia, so the archaeological record has its limitations. From
the Indigenous Aryan side, the references to the Sarasvatl have been produced in sup¬
port of a greater antiquity for the text through this philoarchaeological method as have
the less convincing references to the fire altars (although these too can be challenged).

However, there is a more important datable item that is first mentioned in the
Atharvaveda (11.3.7; 9.5.4) and again in the Satapatha Brahmana: krsna ayas, black
metal/bronze’ namely, iron. 5 Smelted iron first appears in the archaeological record in
a variety of places by the thirteenth to tenth century B.C.E.— including the Deccan
(Chakrabarti [1997a] notes that the iron in inner India is attested earlier than in the
northwestern borderlands). The mention of iron in these texts is as solid a chronologi-

The Date of the Veda 247

cal indicator as one can expect in the reconstruction of protohistory and cannot be
cursorily dismissed. Moreover, the dates for iron in India are in sync with the dates for
this metal attested in central Asia and Iran: if anything, the Indian context is the earli¬
est and most likely to have influenced the others (Koshelenko 1986, 73). Even if, with
Dikshit (1985), we push this back to a highest possible terminus a quo bracket of 1500
B.C.E., we have a significant terminus post quo basis for arguing that these texts were
written after the attestation of iron, by the same logic and method, discussed earlier,
that has convinced most IE’ists to require that the Indo-Europeans were a united entity
during the Copper Age but before the Bronze Age. One would have difficulty on philo¬
logical grounds, accordingly, in placing the Rgveda, too much earlier than the Atharvaveda,
since the language of this text, although later, is not sufficiently different to warrant an
interval of too many centuries. The iron evidence supports the consensus that will place
the date of the Rgveda somewhere within a 1900-1200 b.c.e. bracket.

This bracket seems to be justified, provided we can be assured that the krsna ayas of
the texts refers to smelted iron objects and not to iron ore. After all, krsna ayas simply
means ‘black metal.’ As has been discussed previously, while there is no evidence of
smelted iron in Harappa, iron ore has been found in eight sites, and household items
have been made from it. In Mundigak, five iron items dated between 2600 and 2100
b.c.e . 6 were found, including a copper/bronze bell with an iron clapper, two iron “but¬
tons” on a copper/bronze rod, an iron button on a copper/bronze mirror, and two
indistinct lumps of “carbonates of iron” (Possehl 1999b, 159). Some of these seem to
be items of everyday use. Ocher sites have revealed: Said Qala Tepe, “ferrous lumps’
(2700-2300 b.c.e.); Ahar two iron arrow heads (ca. 1275 b.c.e.); Chanhu-Daro, an
“iron artifact” (context questionable); Mohenjo-Daro, some lollingite (an iron bearing
mineral that may have been used in copper smelting); Lothal, a fragmentary piece of
metal (2500-1800 b.c.e.); and Katelai Graveyard in the Swat Valley, a single piece of
iron (1500-1800).

In actual fact, it has even yet to be discounted that some of these might have been
even smelted: “None has been analyzed to determine their technical properties and we
do not known which of them is meteroic and which (if any) were smelted” (Possehl,
1999b). Either way, items made of black metal go back to the Bronze Age in Harappa,
whether or not they were smelted. This does somewhat minimize the persuasiveness of
the 1100 b.c.e. date for the Atharvaveda on the ground that it refers to krsna ayas. The
black metal could have been accidentally encountered as a by-product of the smelting of
copper, manipulated in some of the ways noted here, and referred to as the ‘black’ ayas.
Moreover, since the Rgveda knows no iron, one should not be surprised if the listed
items surface in Indigenous discourse as proof that the Rgvedic Aryans must have ex¬
isted in the area before such awareness of iron since their texts do not mention this
metal.

Moreover, 1 have encountered the argument that we must consider the possibility
that the word krsna ‘black’ was inserted in the texts at a later period to qualify the term
for bronze. The rationale here is that the older versions of these texts may have origi¬
nally contained simply the word ayas, meaning metal or bronze, and krsna was added as
a supplement to the original wording in later redactions of the text after krsna ayas ‘black
bronze’ had been discovered and was becoming a more commonly available metal. The
original recension of the text, however, could have been handed down through the cen-

2 48 The Quest for the Origin of Vedic Culture

tunes from a period much older than this discovery in 1100 b.c.e., at which time the
metal referred to by the text could have been bronze. Although this possibility might
appear to be a case of special pleading, it must be kept in mind. After all, it has been
argued (e.g., Dolgopolsky 1990-93) that *ayes originally simply referred to metal in Proto-
Indo-European times, and that the word was transferred to bronze at a later point when
this metal became widely utilized by the Indo-Europeans. Such semantic transferrals are
quite common. So it is possibile that the older texts knew only bronze, which was then
modified to iron in the course of oral transmission. This would simply involve adding
something rather titan changing an existing word in a sacred text.

This explanation, however, which might work for the prose Satapatha Brahmana,
does not work so well for the Atharvaveda, which is in meter. Any inserted word would
disrupt metered verse. Since I have noted in the conclusions to previous chapters that
the Indigenous Aryan position can remain cogent only if the Indo-Aryan language can
be argued to have had a great antiquity, if a date shortly prior to the discovery of smelted
iron is proved to be indisputable, the Indigenous Aryan argument loses cogency. There¬
fore, the iron evidence must be questioned by those supporting this point of view.

Chakrabarti (who has written extensively on iron in both literary and archaeological
contexts) argues (1986) that one should be wary of “underlining a particular set of ar¬
chaeological data and arguing that these data conform to a particular section of the Vedic
literary corpus without at all trying to determine how this hypothesis will affect the other
sets of the contemporary archaeological data and other sections of the Vedic corpus”
(74). If iron is to be extracted as a chronological indicator, it must be juxtaposed with
other indicators. Chakrabarti produces another reference from the same text in an at¬
tempt to counter the implications of the iron evidence:

Another instance which comes readily to mind is the reference in the Satapatha Brahmana
to the spread of agriculture in the Sadanira or Gandak river valley. . . . agriculture was
well established in the Gandak valley as early as the third millennium B.c. The SB tradi¬
tion apparently contains a dim protohistoric memory. To fix this text within the straight-
jacket of a late date (c. 700 B.c.) is surely not a logical exercise. (74)

Chakrabarti does not state his reference, but he presumably is referring to the same
verse quoted in chapter 3, which describes the first arrival of agnt, fire, on the other
side of the Sadanira River. Chakrabarti seems to be arguing that agriculture requires
knowledge of fire to prepare the land: if agriculture was on the other side of this river
in the third millennium b.c.e., then so must fire have been. Accordingly, the Satapatha
Brahmana must date back to at least the third millennium b.c.e.

In terms of argumentum ex silentio, Sethna (1981) has written a whole book predi¬
cated on the fact that since the Vedas do not mention cotton, and cotton is known
from Harappan times onward, then the Vedas must be pre-Harappan. 7 Southworth
(1988, 663) uses the same lacuna to draw the opposite conclusion, namely, that the fact
that the earliest Vedic texts do not mention cotton—nor, for that matter, wheat, dates,
sesamum, and rice, all of which are present from Harappan times onward—suggests
that the early Indo-Aryans were unaware of such things and so must be post-Harappan.
Philoarchaeology, then, like everything else, can be used to support different conten¬
tions depending on different presuppositions and cannot easily solve the problem of

The Dace of the Veda 249

Vedic chronology to everyone’s satisfaction (especially since the texts are not catalogues
of agricultural products).

Ultimately, all that can be authoritatively established about the chronology of the
Vedic corpus (viz., Sarhhitas, Brahmanas, Aranyakas, and the earliest Upanisads) is that
it preceded the Buddhist literature that refers to it. Such a terminus ad quern seems
reasonable. But one of the main rationales offered for establishing an initial date for
the composition of the Rgveda is the entrance of the Indo-Aryans themselves into the
subcontinent, since “the determination of the terminus a quo is closely connected with
. . . the vexed problem of the time at which the Aryans arrived in India” (Gonda, 1975,
20). And this is clearly not accepted as an a priori fact by the subjects of this study.
Since conventional scholarship has assigned the Aryan arrival in India to about 1500
b.c.e., Max Miiller’s date of 1200 b.c.e. for the composition of the Rgveda remains
acceptable to most Indologists. This allows an interval of three hundred years for the
Aryans to settle down in the Panjab and completely forget about their overland saga.
Needless to say, since the whole Aryan arrival is questioned by the Indigenous Aryan
school, any date for the Veda predicated on proposed dates for supposed Indo-Aryan
movements is, by extension, considered unacceptable.

The Avesta, as we have seen, anchors its chronology on the same Indo-Aryan migra¬
tions, so it is in no position to otter any extraneous assistance. Moreover, as Gonda
(1975) notes, “besides the uncertain date of the Avesta, the cases of cultural, stylistic
and lexicographical parallelism between texts of this description do not necessarily point
to simultaneity” (21). The same can be said for the Mitanni treaty, generally dated to
the sixteenth century b.c.e. Here we do have archaeologically datable evidence for the
Indo-Aryan language. The Mitanni treaty' provides additional evidence that fits smoothly
with a date of about 1500 b.c.e. for the Veda in India: some Indo-Aryans settled in the
Near East shortly before, or while, others were settling in India. However, the same
arguments could be raised here: the parallels between the Mitanni documents and that
of the Veda also do not necessarily demonstrate simultaneity. The case of modern
Lithuanian, which has preserved very archaic Indo-European features into the modern
period despite being separated from the protolanguage by so many millennia, demon¬
strates that the Mitanni could, likewise, have preserved Vedic forms that may have been
much more ancient than the second half of the second millennium b.c.e. And Misra,
(1992) we may recall, holds that the language revealed in the treaty contains Middle
Indo-Aryan forms and could therefore be much later than the Rgveda. Of course, this
view is by no means a fait accompli, there are other ways of interpreting the evidence,
but it cannot be rejected as a possibility. Unfortunately, we are dealing with a few words
in these documents, which does not allow us to document the full degree of correspon¬
dence between the Indo-Aryan of the Mitanni and that of the Rgveda. 8 Those arguing
for a greater antiquity for the Vedic texts will in all events have to argue that the Indo-
Aryan language maintained unusual linguistic stability over a vast period.

Woolner (1986, 80) provides some parallels for this: he notes that in Egyptian records,
the lapse of a thousand years made little difference to language and style, the language
of King Sargon in the Assyrian records appears to be much the same as that of Nebuchad¬
nezzar about two millennia later, and even Chinese has changed relatively slowly dur¬
ing the last two thousand years, apart from phonetic decay. He acknowledges however

250 The Quest for the Origin of Vedic Culture

that this may be representative of the stability of the script rather than the underlying
phonetic system, or it may indicate a fixed literary language, which could be expected to
remain more stable than a folk one. Nonetheless, Woolner is clearly unconvinced by
the assumptions underpinning the date commonly assigned to the Vedas and attempts
to undermine this by means of analyzing the often-cited claim, first suggested by Geldner,
that the difference between the Avestan and Vedic dialects is no more than that be¬
tween any two Romance languages. 9

While pointing out a variety of assumptions and pitfalls involved in his exercise,
Woolner nonetheless proceeds to compare Spanish and Italian versions of the psalms
on the grounds that such hymns provide some sort of parallel to the liturgico-poetic
nature of the two Aryan texts, the Rgveda and the Avesta. In addition, these two lan¬
guages are commonly held to be as approximately divergent as their Aryan counterparts. 10
He dates these psalms to a mean of about 1500 c.e . 11 Adopting Woolner’s method,
but adjusting the dates he assigns to the Veda, the Avesta and the joint Indo-Iranian
period to reflect the more recent opinion of specialists (which will produce results that
illustrate his point even more effectively than the dates he adopts), we will follow Boyce
in dating the upper limits of the oldest parts of the Avesta to 1500 b.c.e., the acceptable
date of 1500 b.c.e. for the Veda (Witzel’s earlier date of 1700 BCE works just as well, as
does a later one of 1200 b.c.e. for either text) and assign a (relatively) uncontroversial
date of about 2000 bce for the common Indo-Iranian period. This produces a period of
about five hundred years drat seems to represent, by communi consensu, the time taken
for the two dialects to split from their common origin.

The point is that upon applying this five-hundred-year period to the supposedly par¬
allel case of the Romance languages, one would arrive at a date for the common Latin
period of 1000 c.e. (1500—500), which, as Woolner’s notes, is “patently absurd,” being
about twelve centuries too late! If, Woolner argues, we calculate the period we know
elapsed from the Latinization of Spain (we have the Roman colonization of this area as
a definite historic starting point for the Latin origin of Spanish) until the period when
Spanish and Italian had taken their distinctive forms as represented in the psalms, we
arrive at a figure of seventeen centuries. Applying this known figure of seventeen centu¬
ries to the hypothetical period separating the joint Indo-Iranian from Vedic and Avestan
in 1500 b.c.e. (which is supposed to parallel the Romance situation), one would arrive
at a period of 3200 b.c.e. for this proto-Indo-Iranian period. 12

Obviously, there is much that can be challenged in such an exercise, since there are
so many variables that can be brought into question. Moreover, die method is predi¬
cated on the assumptions of glottochronology, which is no longer accepted by most
linguists, since, as I have noted, languages cannot be demonstrated to evolve at predict¬
able rates. But Woolner’s point, I think, is to show how arbitrary the allotment of any
kind of chronological assignment actually is. Woolner (1986) concludes: “Perhaps it
may be asked—is there then no limit? Can we equally go back to 3000 or even 4000
BCE. 7 ” He concedes that it is doubtful whether anyone would propose so remote a
date as 4000 b.c. for the actual text of any hymn, or for the Aryan settlement in the
Punjab (which he accepts), but argues that “the highest possible date for the Vedic
deities, and of many elements of Vedic culture, not to speak of possible reminiscences
of older periods, is a very different matter” (83). Accepting such dates for the actual
Vedic texts themselves would involve expecting philologers to allow a period of lin-

The Date of the Veda 251

guistic stability in a literary language that is far greater than anything recorded any¬
where else in the world.

Most Indigenous Aryanists seem to feel that there is no convincing reason that this
cannot be the case: after all, the oral recensions of the texts have been maintained with
meticulous precision for at least three millennia, and this has no known parallel any¬
where else in the world. But apart from this, for well over a century, many Indian scholars
(and several Western ones) have been convinced that the Vedic texts do actually con¬
tain solid philological evidence that requires the texts to be dated in the third or even
fourth millennium b.c.e. (or earlier). This evidence is astronomical.

Astronomy and Vedic Chronology

Europeans first became interested in Indian astronomy for the same reasons that they
eagerly scrutinized ancient Indian texts in general: there was a sense of concern regard¬
ing whether Sanskrit sources would discredit or substantiate Old Testament narrative.
As we have seen, Sir William Jones was very preoccupied with the traditional date for
the Kali Yuga. If this could be established, he felt confident that he could assign an
average reign period to the kings of the pre-Kali Yuga dynasties and thereby determine
whether Indian history could be accommodated within the generally accepted date for
the Flood. 13 In 1790, he was the first to attempt to use the astronomical method to
calculate the age of Parasara Muni, whom he supposed might have lived until the Kali
Yuga. Taking statements from Varaha Mihira describing the position of the sun in the
constellations at the equinoxes in Parasara’s time, and estimating the degree of differ¬
ence in its equinoctial position in his own time due to the precession of the equinoxes,
Jones calculated that the Muni must have lived about 1181 b.c.e. The traditional date
for the Kali Yuga, by such reckoning, was a serious miscalculation by the Brahmanas. 14

H. T. Colebrooke (1803) also used this method to calculate the degree of difference
between the constellation in which Spring, and hence the vernal equinox, began in the
Veda and the constellation in which it began in his own time. He concluded that the
Vedas “were not arranged in their present form earlier than the fourteenth century before
the Christian era” (284). Like Jones, he too felt assured that this invalidated the tradi¬
tional date for the Kali Yuga. In his opinion, this information provided a terminus a
quo for the compilation of the Vedas. Not all scholars were so conservative, however.
The astronomer and onetime mayor of Paris, Jean-Sylvain Badly, in his Histoire de
Uastronomie ancienne ex moderne (1805), felt that “these tables of the Brahmana are
perhaps five or six thousand years old” (53; my translation). 15 Badly approved of the
traditional date of the Kali Yuga, and seemed to have convinced at least some of his
colleagues such as Laplace and Playfair of the accuracy of the Indian astronomical claims
(Kay, [1924] 1981, 2). This was bitterly opposed by another astronomer, John Bentley
([1825] 1981), with a concern that we have seen was typical for the times: “If we are to
believe in the antiquity of Hindu books, as he would wish us, then the Mosaic account
is all a fable, or a fiction” (xxvii). As was discussed in chapter 1, much was at stake in
such differences over the antiquity of the Indian sources.

Almost a century later, in 1894, another remarkable controversy erupted briefly in
the pages of Indian Antiquary and other Indological journals concerning sensational claims

252

The Quest for the Origin of Vedic Culture

for determining the dates of Vedic texts. Bal Gangadhar (Lokamanya) Tilak and Herman
Jacobi, completely independently of each other and initially oblivious of each other’s
research, claimed a great antiquity for Vedic culture on the basis of astronomical clues
that were hidden in the Brahmanas and other Vedic texts, and one far greater than
anything that was under consideration in the academic circles of their day. 16 The two
scholars submitted their results to Btihler for his consideration within six weeks of each
other. After their findings had been made public, Biihler published an article of his
own, which approved of their basic conclusions and contributed additional material in
support of their arguments. These publications were followed in 1895 by three responses,
from W. D. Whitney (published posthumously), G. Thibaut, and H. Oldenberg—all of
which attempted to refute the arguments of Jacobi and Tilak.

After this sudden and quite dramatic interchange, and although Jacobi continued to
defend his contentions, (see Jacobi 1895, 1909, 1910), the astrochronological method
disappeared from the pages of Indological journals—at least in mainstream academic
circles in the West. Tilak, unbowed, went on to use his findings in his book The Arctic
Home of the Aryans. As is obvious from the title of this book, Tilak was not attempting
to assert that India was the homeland of the Aryans—he accepted, at least at this point,
that the Aryans had invaded India—but was arguing that the Vedic texts might be much
older than had been generally acknowledged. 1 ' He had also begun to prepare another
book on the subject before his demise, of which the introduction, the first chapter, the
outlines of eight more chapters, and some incomplete appendixes were published post¬
humously under the title Vedic Chronology and Vedanga Jyotisha. Since then, numerous
Indian scholars have continued to insist that the method and its conclusions are valid.
In 1965, N. N. Law, founder-editor of the Indian Historical Quarterly, resurrected the
whole debate in support of the positions of Jacobi and Tilak, and die “astronomical
evidence” has kept surfacing ever since. As Pingree (1970) puts it, “speculations abound
concerning alleged astronomical data in the Vedas, in the Brahmanas, and in other early
Sanskrit Texts” (534).

The astronomical data are not dismissed quite so lightly by most Indian specialists
in this field. 18 The Indian National Science Academy of New Delhi, for example, pub¬
lished a History of Astronomy in India in 1985, wherein the Indus Valley and the Brahmana
period are correlated, and dates in the seventh millennium b.c.e. for the Vedas are
proposed. In 1994, a number of the papers at a conference at the B. M. Birla Science
Center and Planetarium in Hyderabad on “Ancient Indian Chronology” were based
on astronomical claims. Such assertions have a direct bearing on the age of the Rgveda
and, hence, on the greater issue of Indo-Aryan origins. The Rgveda is located primarily
in the Punjab. If these claims of enormous antiquity for the texts have any validity, then
the case for the Indo-Aryans being placed in this area at a time early enough to super¬
sede any attempt being made to trace them archaeologically or even to decipher the Indus
script becomes much more feasible. Accordingly, since die astronomical debate is clearly
not settled in the minds of many scholars, it seems useful and relevant to examine the
more sober and carefully presented arguments connected with this method.

The main issues involved here are actually not astronomical; that is to say, the con¬
troversy is not one involving the actual astronomical computations, which are quite
elementary and not under dispute. As Tilak remarks, the debate is primarily exegetical
and to be judicated by Indologists who need not feel they must defer to astronomers. 19

The Date of the Veda 253

The controversy is over the interpretation of certain passages in Vedic texts. The only
technical astronomical knowledge required to evaluate this method is that of the preces-
sion of the equinoxes. The rate of precession is best calculated on the basis of any of
four conspicuous days that occur during the course of the sun around the heavens: the
two solstices (summer and winter) and the two equinoxes (vernal and autumnal). 20
Because the earth wobbles on its axis like a slowly spinning top, solar events located
upon the celestial equator, 21 such as the equinoxes and solstices, also slowly drift among
the stars in the celestial sphere. This westward drift is called the precession of the
equinoxes.

Since the Indie system is moon-based, the celestial sphere is divided into twenty-
seven naksatras ‘lunar mansions’ or constellations in Indian astronomy (Western
astronomers have divided the celestial sphere into the twelve constellations, since West¬
ern astronomy is based on the movements of the sun). It takes 27.3 days for the moon
to make a complete revolution in the sky, so the celestial sphere was divided into twenty-
seven portions such that the moon could appear in a different constellation each night.
It takes approximately twenty-six thousand years for a point on the celestial equator to
make a full circuit of the celestial sphere as the result of precession. Since the Indian
naksatras are twenty-seven in number, points such as the equinoxes or solstices will be
situated in a particular naksatra for about a thousand years before slowly moving into
the adjacent naksatra. In a period overlapping about 3000 b.c.e., the sun would have
been in the naksatra constellation of Mrgasira at the vernal equinox; around 2000 b.c.e.,
in Krttikd; in about 1000 b.c.e. in AsVini, and so on to the present day. Obviously,
such a phenomenon can be an invaluable tool for dating ancient texts or inscriptions,
provided they contain unambiguous information about the position of the sun or full
moon in the zodiac at the equinoctial or solstitial points and provided these points have
been calculated accurately.

Our knowledge of Vedic astronomy 22 is gleaned mostly from peripheral statements—
usually connected with the times prescribed for performing sacrifice. Although there
were professional naksatra darsas ‘observers of the lunar mansions’ at the time of the
Brahmanas (see Vajasaneyi Samhita, xxx, 10; TaittirTya Brahmana, iii, 4, 4, 1 ), the
oldest astronomical manual preserved for posterity is the Vedangajyotisa, whose date
will be discussed later. In terms of specific astronomical information enunciated in the
Vedic texts that is relevant to this debate, there is a year of 360 days mentioned in the
Rgveda, called samvatsara, which was divided into twelve months. 23 The months were
divided into two paksas: puwapaksa, between new moon and full moon, and aparapaksa,
between full moon and new moon. The year was divided into two ayanas: the uttarayana ,
the six months when the sun travels north from its course nearest the horizon—from
winter solstice to summer solstice—and the daksinayana, or pitryana, when it travels back
down to the south. There is no explicit mention of solstices or equinoxes, although
knowledge of the solstices, at least, is implicit in any awareness of ayanas. There are
twenty-seven or twenty-eight naksatras, which are always listed in the texts as beginning
with Krittikd (several of which are mentioned in the Rgveda). The first explicit reference
to the beginning of the year, which is the crucial issue in this method, occurs in the
Vedangajyotisa, which states that the new year begins at the winter solstice. In the his¬
torical period there were several year beginnings, and both Jacobi and Tilak argue that
the same was also the case in the Vedic period. 24

254 The Quest for the Origin of Vedic Culture

I would like to reassure the reader at the outset that this section does not involve
technical astronomical calculations, although there is an unavoidable plethora of naksatra
names occurring in close juxtaposition that can be difficult to keep track of. The confu¬
sion is exacerbated by the need to remember whether the sun, the moon, or the month
is being referred to, whether these are in conjunction with one of the solstices or one
of the equinoxes, and which epoch in time is being discussed; keeping track of this
material can be frustrating. I have attempted to spell everything out as clearly as pos¬
sible to minimize confusion, and have bypassed the technicalities of Jacobi’s, Tilak’s,
and Thibaut’s interpretations of the astronomical passages, along with N. N. Law’s
counterresponses for which the reader can best refer to the originals. I will give only a
brief summary of some of these arguments in this section, focusing primarily on the
different underlying assumptions involved in the debate. It will be helpful to refer to
table 12.1 for the next section.

Beginning with the more speculative and controversial claims, Jacobi (1909) com¬
bining the interpretations of two separate hymns, found reason to suppose that at the
time of the Rgveda, one of the dates for beginning the year was at the commencement
of the rainy season—which corresponds with the summer solstice—with the sun in
Phalgunl. 25 This would have occurred sometime between 4500 and 2500 b.c.e. (see table).
He bolsters this claim by finding other statements from various sutras which prescribe
the commencement of the updkarana, the period dedicated to study, for the rainy sea¬
son (and, hence, the summer solstice) with the full moon in Bhadrapada. As can be
seen from the table, the full moon was in Bhadrapada when the sun was in Phalgunl
during this 4500-2500 b.c.e. period.

However, Jacobi found other sutras that prescribed the full moon of Sravand for the
same event (which the table chart indicates occurred in the period 2500-600 b.c.e.). Jacobi
explains this discrepancy by claiming that the former prescription occurred during an
earlier period, but, in the course of a millennium or more, due to the precession of the
equinoxes, the summer solstice no longer coincided with the full moon in Bhadrapada
but with that in Sravand instead. Out of deference to the sanctity of the Veda, the out¬
dated prescription was not removed from some of the texts. Buhler, in support of Jacobi’s
(and Tilak’s) line of reasoning, found references in eighteen different sutras assigning
either one or the other, or both, of these full moons as suitable for the event. He, too,
felt that the older, outdated prescriptions were sometimes replaced, and sometimes kept
or juxtaposed alongside the more current prescriptions, out of deference to the sanctity
of the texts.

Table 1 2.1 The processional chart of Naksatras.

Time of year

4500-2500 b.c.e.

2500-

600 B.C.E.

Position
of sun

Position
of full moon

Position
of sun

Position
of full moon

Vernal equinox

Mrgasira

Mula

Krttika

Visakha

Summer solstice

Phalgunl

Bhadrapada

Magha

Sravana

Autumn equinox

Mula

Mrgasira

Visakha

Krttika

Winter solstice

Bhadrapada

Phalgunl

Sravana

Magha

The Date of the Veda 255

Tilak arrived at identical conclusions, although mostly supported by different pas¬
sages and, unaware of Jacobi’s work at the time of his research. Coincidentally, both
Jacobi and Tilak did note in harmony that the etymological meanings of the naksatras
make perfect sense when connected with the various year beginnings that they have
postulated for this 4500-2500 b.c.e. epoch (i.e., Mula ‘root’ and its older name Vicrtau
‘the dividers’, both terms appropriate for the first month of the year, as is the term for
the preceding month—the last month of the previous year—called Jyestha 'the oldest’).
Like Jacobi, Tilak, also finds support for a year beginning with the summer solstice in
this ancient period. Among a variety of arguments, he notes that the pitrydna, the fort¬
night dedicated to the forefathers, is prescribed for the two weeks after the full moon in
Bhadrapada. He wonders why this date was selected and argues that this makes perfect
sense sometime during the 4500-2500 b.c.e. period. At this time, this two-week period
would have occurred right after the summer solstice (with the sun in Phalguni ) and would
then have been the first two weeks of the pitrydna, the six months of the year dedicated
to the pitrs, when the sun begins its journey ( ayana ) southward. Situating the two-week
ceremony to the pitrs at the beginning of the pitrydna made logical and coherent sense
to him. Nowadays, due to precession, the fortnight of the pitrs no longer corresponds to
the first two weeks of pitrydna. Tilak held that there is no other logical reason or expla¬
nation for its present occurrence and observation in the middle of the pitrydna.

Like Jacobi and Buhler, Tilak also argues that these ancient prescriptions were main¬
tained in some texts out of deference to the sanctity of the sacred texts and juxtaposed
with other dates that were inserted in later periods to correspond more accurately to
later astronomical situations. Tilak produces a variety of passages that he believes sup¬
port the later epoch with the year beginning with the sun in Krttikd (Pleiades) at the
vernal equinox in around 2500 b.c.e.. For example, he notes that the earlier texts often
refer to the Krttikd as the beginning, or mukharn ‘mouth’ ot the year. Tilak considers
that this beginning must have once occurred a

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Books

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "2001 -The Quest For The Origins Of Vedic Culture The Indo Aryan Migration Debate"

See other formats